John Carmack Leaves Meta | Hacker News Reader

John Carmack Leaves Meta

facebook.com

1,875 pointsviburnum1,169 comments

We changed the URL from https://www.businessinsider.com/john-carmack-meta-consulting... to the letter that everyone is responding to. A copy is at https://pastebin.com/Jk5TrGch (via https://news.ycombinator.com/item?id=34030288 - thanks!)

”It has been a struggle for me. I have a voice at the highest levels here, so it feels like I should be able to move things, but I'm evidently not persuasive enough. A good fraction of the things I complain about eventually turn my way after a year or two passes and evidence piles up, but I have never been able to kill stupid things before they cause damage, or set a direction and have a team actually stick to it. I think my influence at the margins has been positive, but it has never been a prime mover.”

It is simply stunning that the seasoned direction and counsel that someone of John Carmack’s caliber is capable of delivering was not being followed.

Throwaway because it could be easy to identify my position from my normal account name.

Carmack is many thing, engineering genius above them. However, he would frequently wade into areas where he had no experience, demand others do what he said, ignore evidence he was wrong, bully people, and disparage entire teams who were doing good, and in some cases legally required, work. When data proved his idea was wrong, he would say words to the effect of "I don't care, because I still believe I'm right from an ideological background". He would devalue people, there expertise, there experience, and there thoughts because "I'm John Carmack". Truthfully, I have never worked with someone before who was somehow so politely toxic to a workplace.

Carmacks work in VR was absolutely invaluable from a technical standpoint, but VR now is as successful as it is in spite of his influence, not because of it. When I hear people say "If only Meta would let Carmack do what he wants we'd see his ideal VR experience and it would be amazing". You already saw it. It was Oculus Go, and by every metric is was a commercial, financial, and technical, disaster.

Yeah well, you're anonymous, so you might as well be someone with a huge ego who got their idea shut down by Carmack once and is still bitter about it. More often than not, when people think they "prove something wrong with data" it's more that they're "taking some data and interpret in a way that fits their standpoint".

Carmack has proven enough times in the past that he's able to deliver, that he can push technology, knows what's possible and what isn't, and can wrap up a product. "Wading into fields he has no experience in" sounds pretty unlikely for VR given his past work. And I wouldn't consider the Oculus Go a failure, more like ahead of it's time and released too early. A prototype of the quest. But I guess now it's easy to claim everything that's bad about the go was Carmack's work and everything good about the quest and quest 2 was someone else's.

I worked with someone who was a bona fide development genius about 10 years ago and he was a CTO and co-founder at the time and considered himself to be able to input into any conversation with people who were domain experts in fields that he wasn't.

I remember him overruling the Head of Design and enforcing Arial to be the typeface used for the corporate brand because it was the typeface was present on more computers than any other in the world. Suffice to say the original typeface was much better.

Self doubt is a really really important trait to have as a leader - don't automatically assume you're right outside of your area of expertise and that team members can come up with good solutions that aren't yours is the only way you can ever really scale.

The CTO ultimately was forced out and became a specialist consultant and that probably suited him.

It's just really weird to me that people are all shocked-Pikachu about Carmack wanting to be the auteur in this scenario and to wield Meta's effectively-infinite resources in the exact ways that he wanted. It's apparent from every interview and article about him that this is how he operates. It seems to have worked well for him in the past, but it was obviously going to be a major culture clash at Meta unless they gave in and let him run the thing.

Honestly no one really cares about VR that’s the problem.

It’s cool in concept but whose going to shell out cash to be tethered to a machine wearing goggles sitting in a chair?

AR has potential but even that is marginally better than alternative solutions.

Also, I don’t know about y’all but I don’t trust Facebook so I don’t trust Meta. They are a data-leach.

We still probably have a decade or more to go with this technology it has to be affordable, lightweight, AR glasses not tied to a company that sells peoples data!

He was alone on his side of the table for the All In One integrated hardware approach that was the Go and became the Quest and, after years, proved to be the right long term direction. Of course that is out of his mouth according to him but, if true, it is the kind of high level strategy setting that separates market leaders from the also ran's. Very different from task level involvement but friction there should be separate from effectiveness at the CTO's desk level (& I have to guess Boz is the one who's quality of life got improved more than specific engineering teams)

His critique resonates with my expectations from large enterprises, and your perception also resonates from what I expect of large enterprises workers. So I'm guessing he's probably not wrong and your experience was probably an example of him hitting social/political roadblocks. In my arguably limited human sized experience, the vast majority of people at Meta-sized companies prefer the safety and comfort of things going slowly and being discussed at length, and being careful to respect all the hoops that need jumping through. And not having their caged shaken too much. This leads to 5% GPU utilization.

Carmack has more than proved himself. !

Guess what buddy.... the entire fing executive management team is wading into areas they don't know this about, stabbing each other in the back, slowing progress due to corporate BS.

And so what if Carmack was 'wading into areas where he had no experience'? (What was he showing up at the quarterly internal financial review and advising on advance tax strategy with offshore account to line exec pockets???? ;

When you have a talent like that the organization makes space for them. It was never a cultural fit from the start.

I don't work with John Carmack, but I had a meeting with him once. He is brilliant, no doubt, but he is exactly like every single engineer with no professional communications training. He struggles to make his points in everyday developer language, when he needs to speak more calculated, more measured, and with significantly more audience empathy. Just like 99% of us technologists. We're an industry of weak communicators, and it is hurting all of our careers.

You could be right about all of this, but it isn't very compelling without more concrete detail.

From the perspective of a neutral observer, what we're getting here is one anonymous person's interpretation of Carmack's behavior, and we don't know anything about this person. So for all we know, it's equally possible that this person is arrogant and narcissistic and takes disagreement as bullying.

Not saying I actually believe this of you, but there are plenty of such people in the tech industry. So I don't know what to think here.

>> When data proved his idea was wrong, he would say words to the effect of "I don't care, because I still believe I'm right from an ideological background". He would devalue people, there expertise, there experience, and there thoughts because "I'm John Carmack"

Projecting are you? You're devaluing someone with way more experience and accomplishment than you. Did it ever occur to you that he was right? I've read a lot of his posts that cover ideology, and I always agree. Ideology is what it takes to go big and play a long game. If the short term stuff does fail (even his) it is appropriate to fall back to ideology to figure out what to do next.

I also never saw anything revolutionary in his work. He's really good at selecting practical and straight forward approaches to real problems. If he says "that's stupid and here's why", it's worth listening to even if you can make it work.

If i had nickel for every time i had been in a conversation where someone “proved they were right” even though they were nowhere near there, ok, i wouldn’t have millions but I’d have somewhere in the $50s which is a lot higher than you’d expect

Maybe Oculus Go was a failure in all those areas because VR will always be a failure in those areas.

FWIW, my interactions with Carmack have shown him to be extremely pragmatic when it comes to tech. Going quickly from “we absolutely shouldn’t do that” to “ok it makes sense in that context”.

> When I hear people say "If only Meta would let Carmack do what he wants we'd see his ideal VR experience and it would be amazing". You already saw it. It was Oculus Go, and by every metric is was a commercial, financial, and technical, disaster.

Hasn't Carmack always been pretty clear about wanting Meta to deliver an ultra-low-cost high-volume headset prioritizing getting them in as many hands as possible? Did Oculus Go deliver this? Because Carmack has seemed to be constantly complaining about this with every germane talk he's given to date...

I know someone who built a solid business around using the Oculus Go to show off kitchen designs to prospective buyers. The fact that the Go was a cheap-ish stand-alone Android device was really valuable for them.

I've watched many Carmack videos over the years and he never complained about the team not following him or people working for him being not good enough, he always praised the work done and explained in very fine details why the decisions where taken one way or another.

When he wanted to remove a pebble from his shoe, he talked extensively of the company decisions, in the higher ups, which is more than fair given his role.

He never struck me as a "Steve Jobs of coding" (probably today Elon Musk?).

I also had several encounters with John Romero and talked a bit about the times at id Software and he never ever hinted that Carmack was problematic in any way.

He's also obviously not a very good politician/sellperson (he can't sell what he hasn't already produced or envisioned) and suffers bureaucracy, like every normal person here that is not a bureaucrat.

Anyway.

Regardless of the truthfulness of what you write, Carmack has always been able to deliver, both in time and as of code quality and maintainability, one way or another, Meta hasn't.

The evidence pile up more against FB/Meta management than against Carmack, moreover I think it's easy to attack the person taking responsibility in person than those hiding in the shadows, Linus suffered the same destiny, but he created Linux and brought it where it is now, the attackers didn't, so maybe Linus was simply right.

Interesting perspective. One ominous anecdote is I've noticed that kids seem to have lost interest in the Oculus / VR. I'm assuming the business model needs to be bootstrapped with games and then move into other areas. I hope it is successful however, the world will be kind of boring if all we ever have is screens.

I'm gonna dismiss this whole story, since it's anonymous, while the person you're criticizing (Carmack) isn't.

First, we have no idea what "evidence" or "data" you allege that disproved Carmack, or how Carmack insisted on being right "from an ideological background" (whether that even happened at all).

Second, Carmack is one of the most data-driven and performance-oriented programmers you can find, and is quick to admit errors where he is wrong; to suggest that he is ideology-first (which, from his Twitter and public statements, has never been the case) over truth in tech matters - is bonkers-insane.

And it's such an easy thing to call out someone you dislike as "toxic", just because you disagree with them. Your statement also makes no sense: the phrase 'politely toxic' is an oxymoron - as when one 'politely disagrees', it suggests reasonable disagreement, not outright difficultness.

Further, "good, and in some cases legally required, work" can be work that have nothing to do with the end goals - it could even be baseline administrative work.

Not sure I’d trust someone who mixes up ‘their’ and ‘there’ to interpret data proving or disproving some strategy. From the looks of the product, I’d say he was probably right on whatever that stuff was.

By reading between the lines of your story and carmack's, my guess is carmack was against facebook user profiling in VR headsets, and you think the sales numbers are sufficient and prove him wrong. If that's the case I'm with Carmack both because "the ideological background" and because the sales argument does not hold : you don't know what the sales would have been (and will be in the future) if facebook did not strong arm users into surrendering their privacy

> It was Oculus Go, and by every metric is was a commercial, financial, and technical, disaster.

It was the second best selling VR headset at that time, only behind PSVR1. The only problem with it was that it came too late. GearVR had been around since 2015 already, Go was essentially just a standalone version of that, but it took until 2018 to release. That was simply to late, as a year later it was replaced by the Quest1 and 3DOF VR was killed.

Another problem with the Go was that Facebook didn't put effort into good 3DOF content. They never released a VR180-3D camera. Never build their own VR video platform. And finally they made Quest2 deliberately incompatible with Go, despite Quest1 still having some compatibility.

There was and still is a lot of unnecessary fracturing going on in the VR space, even among Meta's own headsets. I think the Go line should simple have been continued with a stronger focus on movie instead of gaming. The Nreal Air shows what is possible when going that direction and Meta with an 8 year head start should have an easy time matching that. Instead we only got Quest2, which is much more heavy and low resolution than it needs to be for a movie headset.

Everyone fails.

Carmack has also had a lot of wins. His accomplishments stand on their own.

Perhaps he is not a good people person.

He puts in an astonishing amount of effort on his pursuits or at least he used to. He may expect the same of others, not understanding that the majority of people have different lives and goals.

He might also just be an asshole.

It will be interesting to see what he does next. I think he still has one or two wins left in him.

He isn't alone.

A core problem is that FB culture is way too positive and happy, and hard criticism is received poorly. The politics to get anything done if your name isn't Mark is borderline impossible.

I left a year ago, and I've become beyond happy.

> It is simply stunning that the seasoned direction and counsel that someone of John Carmack’s caliber is capable of delivering was not being followed.

This doesn't actually surprise me in the least bit, and that's not a criticism of either FB/Meta or Carmack. It's simply that after a couple of decades in industry, I now see that effective organizational leaders are exceedingly rare, and I've seen now many super-experts --legends in their fields-- join big companies and not find their footing.

I also don't really buy the "you gotta be good at cutthroat politics." I have plenty of examples across multiple top-tier companies, of senior leadership who were smart, thoughtful, effective... and still reasonable and compassionate.

I think it's simply that it's tough to move people, plain as that. And in Carmack's case, I will wager that his expertise and track record were not enough to get everyone to drop what they're doing and follow his lead. After all, there are many other legends at Meta too. And, there's an abundance of good ideas all fighting for limited mindshare and limited ability to act.

Ironically I felt reassured reading this. "[...] I'm evidently not persuasive enough. A good fraction of the things I complain about eventually turn my way after [...] evidence piles up" especially resonates with me. While I rationally understand it's just corporate bureaucracy / politics, sometimes I still wonder if I were just a bit more capable I'd be more persuasive.

If John fucking Carmack cannot move the bureaucracy this way, then it's folly for me to try the same. I should accept that we're playing checkers not chess.

This is why there are so many leaks from Facebook. My sister is a reporter and she says that people leak when they feel they can't influence the decision of the org, that they are a Cassandra who is being ignored by the organization. That there is so much leaking from Facebook- much of it from seemingly senior and highly respected people- suggests major problems in the decision making process there.

> This was admittedly self-inflicted – I could have moved to Menlo Park after the Oculus acquisition and tried to wage battles with generations of leadership, but I was busy programming, and I assumed I would hate it, be bad at it, and probably lose anyway.

I think we can only speculate about reasons, so everybody is gonna project. Maybe the Facebook culture is uniquely bad. Maybe Carmack isn't an effective leader. Maybe it's just how large corporations work. Maybe he was dealt a bad hand.

Meanwhile, I admire that he had the guts to see the issue, say it out loud in a self-critical way, and call it quits. Most people don't.

Success in the corporate world requires more than technical prowess. One of those things is influence. It's not for everyone.

I have a genuine question about what to do when one finds oneself in this type of position and one is surrounded by people who seemingly do give a damn (and one isn't compelled to just give up and leave).

In my experience, the software industry, in it’s aggressive desire to be egalitarian, has an authority problem. I have encountered the “I told you so one year ago but we just had to learn the hard way didn't we” situation more times than it seems efficient. I am not unaware that I may be selectively remembering things and I’m not so prideful to ignore that I’ve also made calls that turn out to need adjusting down the road… but it baffles me why people put in positions of power and leadership are so often ignored almost on the whim of some well meaning but misguided other party. Certainly Carmack has the industry respect to be taken seriously. Why wasn’t he?

Where did this come from? This quote only resolves to this post.

Edit: A little more searching and I found the full memo

https://www.businessinsider.com/meta-john-carmack-scathing-e...

That saying about how teenagers and twenty year old think they're immortal? There's some variant of that for programmers. That failure mode is something that happens to other people, to suckers, so I don't have to change direction because some guy who's going grey at the temples tells me I'm walking toward a cliff.

I've been in places where my job became a bit of a cleaner, for things that shouldn't have needed to be cleaned in the first place. When I realize that's what I'm doing - when it's most or all of what I'm doing - I leave. Sounds like John might have similar boundaries, and found himself standing at the edge of one.

I'd really like to see him try his own thing again, and either not sell it this time or be quite a bit more precious about what he's willing to part with and for how many zeroes.

> It is simply stunning that the seasoned direction and counsel that someone of John Carmack’s caliber is capable of delivering was not being followed.

I'd be very careful about making that assertion when you're only hearing one side of the story. There is a good quote from the Dowager Countess in Downton Abbey about why she never takes sides in a broken marriage: "Because however much the couple may strive to be honest, no one is ever in possession of the facts."

I strongly think that applies here.

>It is simply stunning that the seasoned direction and counsel that someone of John Carmack’s caliber is capable of delivering was not being followed.

He worked for an Ad company that voraciously mined through their user's data. Unless he was able to find new ways to monetize their users or bypass iOS's opt in app tracking, I see no reason why they would even care.

I would have been stunned if they had listened to someone wise and decent. They listen to quarterly reports I guess?

I so wish I could "kill stupid things before they cause damage", "set direction" and have people follow, but it seems especially difficult for technical people.

We spend all-day every-day talking to our subservient (a compiler) but a human is fuzzy and unreliable which is compounded by team size. Carmack can realize something in an instant but it is amazing how long it can take an entire team.

One thing I've noticed that really does not work is "complaining", if Carmack thinks he is complaining, then magnify that by 10 for others on the team.

Rather than attacking an idea directly, I'm wondering if ignoring the idea may be more effective. This way there's no one to fight and the idea can slip away without any ego or drama, there's no face to lose.

If you really can't help the situation, then leaving is the last resort. Oh well, on to the next thing, very curious to see what Carmack gets into next.

The bigger the ship the smaller your rudder is.

That's what it looks like from the outside, but the reality is that he works alongside people at the same level as him as far as the company is concerned. His experience is incredible but its likely that the people he works with are also deeply experienced engineering managers too. He's just more famous.

That's the norm these days I think. No one is supposed to be an expert, everyone is equal, everyone's voice matters.

If you are an expert and demand high code quality, you are a class traitor and the bureaucrats will come after you.

This shows in the code quality of the OSS examples I have seen from Facebook.

If Carmack suggested changes to my code I'd implement them in a heartbeat. If he suggested changes to my business I'd be a bit more cautious.

I don't work at Meta, but I assume that, like any other bureaucracy

It's a large company, with home-grown political feifdoms - and technical merits for ideas aren't good enough to get things your way. Even if he is a big cheese in the world of actually knowing things, that doesn't really matter to his executive peers.

You need to both know things, and be very good at politicking to be successful.

Perhaps he hated bureaucrats meddling with engineering and user experience.

Maybe he’s condescending in a team of bright people, and sees the world as him and then everyone lesser?

It’s hard to say what blend of reasons, other than what he explains.

Damn this rings home.

(Im not comparing my skillset to Carmack, im my eyes hes a god of programming and has an insanely deep skillset and knowledge about so much software related)

Sometimes i feel the same, a new feature is requested and i directly see this is an anti-feature, or just overall a bad decision. I voice my concern, but the client request goes first.

Next up the feature is poorly implemented, usually "just like the client wanted it" without any bigger design. Testing is barely done, and finally it goes to prod. Next is a 3 month period of bugfixes, and ultimately this one feature can take up 60-70% of a devs time.

Then the process repeats.

Can people not see the big flashing warning lights of his attitude right there in his faux self deprecating comment? When one talks about killing 'stupid things', that's actually a colleague's initiative, or possibly some project in flight, with a team of people working on it. People putting in blood sweat and tears. Not everyone in the organisation gets to choose what they are working on. For us adults in the room, who have worked our entire careers treating our colleagues with respect, this is not how you effect change in an organisation. Sure, privately one may think this or that is stupid, but you don't go around using that language. It just offsides people. To bring change you need to bring people with you.

I mean the guy came up in the gaming industry, which by many reports is somewhere on the spectrum between techbro dominated monoculture (not unlike this forum), through to just abusive of its employees, so his bar is very low. And yes I know he was brilliant, but there's no special rules for him because he made Quake run fast.

At the level he was brought on, something more than technical skills were required. He failed.

I'm as shocked as you are. What a wasted opportunity by petty people with big egos. And I paid for it literally. I bought a Oculus set and stopped using soon after. What a disappointment. I wanted to support the technology, but my good will was squandered. At least I feel vindicated by every thing John Carmack says because I thought some similar things.

This would surprise nobody who has any experience behind the FB curtain.

Idk I feel like Carmack is a genius, and has done genius tech related things at id. But does he really have credentials to speak on latest modern developments in technology?

Does he have any skin in the game? Last I heard of him was Megatexturing in idtech 5 and that wasn't really a huge marvel nor necessarily did it make the engine the forefront of game tech.

It's interesting to me that, even here where people pride themselves on their intellectual and rational superiority, we can still observe the primitive human tendency to appeal to authority; see the top throwaway comment [1].

[1] https://news.ycombinator.com/item?id=34026253

yeah if Carmack quits on you it's a political problem and not a tech problem.

Where is that quotation from? I don't see that in the linked article.

Seems like Meta is not a good fit for John Carmack.

It's only stunning because programmers view him as a god, when in reality, he is a just a fallible human being. Just goes to show even programmers are susceptible to idolism.

The signs were all there. He entered a den where he doesn't hold founder status anymore, and that the value of his contributions to the company does not hold as much weight.

Where are you quoting this from? A link would be nice, I don't see this in the article.

He might be considered by others as "boomer" that hasn't been relevant "since Quake III"

Meta itself only exists because of poor direction. It is an aberration of humanity. If good counsel and advice were important to Meta, it would shut down on itself because its very existence is an offence to common sense.

"I have never been able to kill stupid things before they cause damage"

One of the most important things an organization can have is a "no man". A project that I was involved in when I worked at Google was completed on time primarily because we had a very senior (DE level) engineer tangentially involved in the project. He was near the end of his career, and he just didn't give a f* about politics. He'd sit in design reviews and rip stupid features to shreds, with accurate estimates of what they'd cost in terms of headcount and project delays. He was probably the most valuable member of the team because he was respected enough that his objections kept the project focused and on scope, and it was a 20% project for him.

The part about not giving a fuck is crucial. You don't get rewarded for preventing bad things.

> You don't get rewarded for preventing bad things.

A few years ago I saw a promotion announcement e-mail come into my inbox for a colleague who sat a few steps from my desk. It was filled with the usual "did this, did that, made an impact, etc." statements but it also had a large section dedicated to the analysis this employee performed and presented to kill a huge initiative before the organization rolled it out. The initiative was very innovative but it ultimately wouldn't have achieved its goals. It was encouraging to see this in a promotion announcement and indicated to me that some organizations do explicitly reward for preventing bad things.

This was at AWS.

You can, but only in the long run, and only if you’re saying “yes” to good things too. We’ve all probably worked with someone who just says no to everything and manages to kill both good ideas and bad. That leads to stagnation, which can be even worse than trying and failing at a bunch of different ideas.

I got top marks in 2021 based on feedback from two high level managers because I "prevented a year of wasted engineer with with a comment on a Google doc". Those marks lead to a nice pay bump.

It really depends on the size of the org and the team dynamics.

In a small startup with a fairly high degree of trust and emotional safety your CEO is more likely to notice that you prevented damage/loss by saying no.

A large org where so much is lost due to the telefone game, misaligned goals, hierarchy and increased complexity is not an environment where this is (usually) noticed or appreciated.

That's generally the case. It's soul-sucking to work when you can't prevent idiocy and at most places you're punished for pushing back hard, so the only response (other than, you know, burn-out) is to leave. When you see senior people you respect getting the hell out for no obvious reason, that's usually a sign that you should be polishing your resume. Bad times are coming.

> You don't get rewarded for preventing bad things.

Akin to not being rewarded for deleting lines of code.*

* unless the deletion measurably improves performance and you get rewarded for the improvement

Like with politics. Everyone notices leader visiting places after disaster while no-one can notice if one was prevented.

In fact, you might feel like a hero for needing to put in OT to put out a garbage fire! Never mind that it shouldn't have happened in the first place...

> You don't get rewarded for preventing bad things.

This is the source of so many problems at large companies. The employees part of the day to the day understand how much of a positive impact this has but the people in charge of the cash don't see that.

The squeaky wheel gets the grease, or the new product launch gets the promotion.

I’ve also seen morale being slowly killed by people who _thought_ they were doing this, but were _actually_ just argumentative defeatists.

I don’t think Carmack - or the engineer you describe - is likely to be one of these - but anyone hoping to emulate them should be conscious of what they’re doing.

> I’ve also seen morale being slowly killed by people who _thought_ they were doing this, but were _actually_ just argumentative defeatists.

Yes, I've also seen this as well. At some point you need people who will go off on a limb and do something crazy. I've seen in multiple times in my career. In particular I remember an architect who'd ruthlessly attempt to take scope out of projects but while we was in charge, we were nearly never innovating. Just delivering features on time and budget. Existing customers were always delighted but we had a hard time reaching new markets.

After he left, there's been a fair bit of projects that "didn't work" but there's been a couple of absolute home runs that would've never gotten off the ground because they were pretty wild.

There's I've heard a few times that resonates with me and I reiterate to all senior engineers I work with when they get into conflicts. "It takes all kinds of people to make a great company" You need dreamers, the people who worry about the money, cold and calculating execute at all costs types and folks and the folks who are happy to toil in a small but critical corner of a massive system to keep the lights on.

In my experience you can get into trouble when you try to play multiple roles at the same time because it's mentally exhausting trying to to do things like sell a vision while being the person trying to keep the economics of the thing from completely falling apart.

This is true. I think a good test for this is that the person making the criticisms also provides alternatives that are reasonable and it's evident that they're trying to make things better and not just arguing for argument's sake.

You need people to say no. You should be saying no more than yes in most cases. You have more important, more profitable paths to take.

Definitely depends on the person. The amount of projects I’ve worked on that were dismissed by such people done anyway and invariably became the saving of the team/org/company is laughable. These always took longer as they were under resourced and the people working on it did because of belief not because it was good for their status in the company. The hallmark of these is the last minute surge in resourcing when management figures out they got it wrong. Sometimes I think it’s almost like they bought an option on saving themselves by badly resourcing something at the start.

> The amount of projects I’ve worked on that were dismissed by such people done anyway and invariably became the saving of the team/org/company is laughable.

That's intriguing. I would love some examples of these projects / products.

> The hallmark of these is the last minute surge in resourcing when management figures out they got it wrong.

Got what wrong? That they've heard the yeas and nays and chosen to undertake the project sounds like mang got something right?

Switching between business and academia several times I am really struck by this.

Meetings where the elephant in the room is not mentioned, because nobody wants to loose face. Really bad practice not commented on because (I guess, I am bad at this) some form of "mutually assured destruction". I will not point out your mistakes so you do not point out mine.

That, in case you are wondering, is from business.

I am really pleased to see people commenting here about business experience that go against my experience. And in my little group (of two, we could easily be five) we are straight with each other. But in the wider group? Rank incompetence, and dreadful mistakes that could be corrected if someone would point it out.

The point is technical excellence, good scientific practice, that is not required for being successful at business. It is al about "people skills". I.e. convincing somebody that they got what they were promised when they did not.

The university system in my country (Aotearoa) has been hobbled by people with business fetish. The meme was in the 1990s that "business is the most efficient". I believed it, until I saw the waste, inefficiency, downright lying and dishonesty, aghh I am triggered, that is business.

The same thing happens to some degree in academia (especially in commerce. My academic experience is science in the 1990s and commerce in the 2010s) but only "some degree". I see it at every nearly business meeting I attend.

I said before, I will say again, how pleasing it is to see comments here from people with different experiences than me.

These dumb things emerge because of two forces: first, the promotion game of people trying to get theirs. Second, price's law.

Price's law is what constrains scope, and the boundary between engineers can often be the luck of having the right opportunity. As such, all the stupid things emerge trying to invent scope. I'm guilty of this, but that was the game.

There is no solution except accept that big companies kind of suck and are slow.

Maybe I am unusually lucky. After managing a small team for a year in a FAANG, I am considerably less jaded.

I think it is entirely possible to be truly focused on the customer, to have a vision and a strategy, and to balance working toward this goal with experimentation.

It is of course also very hard work, it’s easy to make mistakes, and nothing is perfect. But I really don’t see that this is all doomed from the beginning just because of bad incentives and structural forces.

This sounds familiar. It's usually when I've decided to resign in a few months because thing don't change that I start not giving a fuck and actually enacting change. I wonder how to break that cycle.

It might sound easy to "just don't give a fuck sooner" but it's exhausting. The constant uphill struggle is not something I want to live with.

One of my favourite board games is called Burn Rate. The premise is a parody of silicon valley during the boom. You compete with the other players for talent (each employee is a card, with varying skill levels).

You could slow down other players by attacking them with “bad idea” cards that would tie up one or more of their engineers for an amount of time. To defend against bad ideas, you need a good manager (one type of employee card).

I could be misremembering some of the details, but the wisdom of the above game mechanic has stuck with me :)

As a counterpoint, this does add to Google's already established reputation of not shipping stuff. "No men" are generally a bad idea, they kill innovation and slow things down.

Big orgs reward stupid BAs for coming up with “exciting looking stuff” and the most engineers gladly take it up their az and all would say yeah we can do this like it would make them look bad if they said otherwise. The ones that speak up would usually get grilled and questioned like their job was on the line.

He just founded a new company called Keen Technologies to work on AGI. Not surprising that he wants to focus on that now. He's been part-time at Meta for years. I'm interested to find out what kind of business model he's planning for an AGI company.

Edit: he posted his leaving message publicly here: https://www.facebook.com/100006735798590/posts/pfbid0iPixEvP...

Additional public comments on Twitter here: https://twitter.com/ID_AA_Carmack/status/1603931901491908610

> As anyone who listens to my unscripted Connect talks knows, I have always been pretty frustrated with how things get done at FB/Meta. Everything necessary for spectacular success is right there, but it doesn't get put together effectively.

> I thought that the "derivative of delivered value" was positive in 2021, but that it turned negative in 2022. There are good reasons to believe that it just edged back into positive territory again, but there is a notable gap between Mark Zuckerberg and I on various strategic issues, so I knew it would be extra frustrating to keep pushing my viewpoint internally. I am all in on building AGI at Keen Technologies now.

@dang can you change the link to one of these?

It's very interesting to me that he's giving up on VR entirely. If he thought it was just an issue with Facebook, presumably he'd jump in somewhere else and get the impact he was looking to have.

That's another sign to me that VR is once again not going anywhere. Or rather, half of it isn't. Oculus-style VR is two best in one: 3D persistent virtual worlds and stereoscopic facehugger interfaces. The former has had great success, mostly in games, but the latter has spent decades as the thing that people are supposed to want but don't actually use much when they get the chance.

It wouldn't surprise me if he has a noncompete in play that prevents him from continuing his VR work for a few years.

I think it's less of a judgement about VR and more of a judgement about AI. If you believe AGI is within reach (as he does and I agree), working on literally anything else seems like a waste of time. It's impossible to overstate the impact it will have.

> That's another sign to me that VR is once again not going anywhere

That's not Carmack's opinion. To quote from his post:

"Despite all the complaints I have about our software, millions of people are still getting value out of it. We have a good product. It is successful, and successful products make the world a better place."

"the fight is still winnable! VR can bring value to most of the people in the world"

The main problem with VR in my opinion is a lack of software. Most of the games produced have been either toys that are little more than a tech demo, or ports of other games not designed around VR. There haven't been many serious attempts to harness VR for non-game playing purposes.

I think the future is still bright for VR. We're currently in a bit of a local hype-cycle trough, but the tech is only going to improve.

>That's another sign to me that VR is once again not going anywhere

Maybe it's just a sign of how hot AI is right now.

Try it yourself. Strap something 8 hours or longer around your head, don't move, because every small tremor distorts the vision of the device.

VR in its current form is eye cocaine, nothing more. Why watch poppy avatars hop around in a virtual reality, while you are bound to do nothing?

In every way, VR should be restricted to very few use cases, not to - hello Zuck - rebuild reality completely and hereby track everything you do or see to deliver even more addictive material to your eye vision.

It sounds cool, some use cases look cool, but the very fact that hardly anyone at Meta uses their own device/creature, speaks volumes. I am glad, JC takes consequences and abandons this experiment.

Or maybe you know he is simply a lot more excited about AI than VR now and that's it. No need to stick with something that doesn't excite you any more just because you were excited about it at one point and spent several years working on it :-)

I’d wait and see how Apple’s device is received before making any long-range pronouncements.

I suspect he's going to get smacked hard by the AGI problem and ultimately concede defeat on his biggest goal while achieving some success in the current less-than-general ML stuff.

If you solve a few more problems with chatgpt it's going to become useful enough that people are going to stop caring so much about the AGI label. It's going to quack like a duck enough to change a lot of industries.

I also think he's going to lose it massively (technical background aside, i.e. he's not an applied mathematician, it's a huge problem) BUT the AI industry is currently run with a very childish approach to software and programming, if his company can attack the infrastructure wisely they could really improve the industry. No idea how to make that money money though.

Have a listen to some interviews. Doing it out of love for the problem more so than anything else. https://www.youtube.com/watch?v=I845O57ZSy4

It's as if nobody's learned anything from the 'self-driving car' boondoggle of the last 20 years or so.

Any problem can seem easy when you don't know the things you don't know.

Honestly seems like a better fit for him than VR. VR may be games, but everyone is saying his strengths are technical, not organizational. Seems like he wants to work on hard technical problems and thats what he's good at.

I believe we're still 15 years out from AGI, but I also think we could start creating autonomous agents right now that make people feel like it's almost AGI. I think Carmack could get something out the door that feels like that.

It’s not exactly a bold predction that someone will fail to solve a grand, millennia-old problem of philosophy and science.

I don’t think you’d need to be alone in solving the AGI problem.

Carmack is an engineer at heart. I am sure there are some yet undiscovered technical blockers in that space that he will solve.

>he's planning for an AGI company.

This is something I think will break the mystique around Carmack. Back in the 1990s and early 2000s he was considered the whiz kid when it came to rendering engines but in reality something that's as nuanced and still not fully understood (intelligence, especially AGI) is likely too far out of his expertise to develop. Why? Because the last 60 or so years of development in the field hasn't yield much in the way of results in pure research or practical application. Machine learning and other hat tricks aren't even in the realm of AI properly but rather probability models that work well enough on a narrow set of problems to possibly commercialize them (ex. chat bots for help desk replacements). The thought that anyone, and I mean anyone, is going to develop artificial general intelligence before the 2050s or even later is to me laughable. We barely can understand how to talk to parts of our own brains, we're still learning more from other so-called lower lifeforms with respect to how they solve problems (fun fact: it seems bumble bees like to play or at least do things that are roughly analogous to being play-like, all without any kind of human or hominid-like brain so fancy that). So, it just seems like hubris to imagine any corporation, research department, or a singular scientist cracking this nut. I think we're more in the technological Bronze Age with respect to computing than anything else. It would be nice we finally got started on our computing Iron Age.

He believes that AGI is an engineering problem because of the vast compute resource required for it. Being an engineering problem he believes he can make an impact.

That would be utterly wrong.

+1 to changing the post to Carmack's published post on Facebook or Twitter without the Business Insider fluff.

If you solve AGI, all business is your plan.

Keen Technologies. That’s a great name.

I think it is based on id Software's Commander Keen[0] game.

[0]: https://en.wikipedia.org/wiki/Commander_Keen

Why would he change the link? We're discussing this article.

“During Meta's developer conference in October, Carmack hosted a solo hour-long talk about the company's Oculus or Quest headset. He admitted he had many things to be ‘grumpy’ about, like the company's rate of progress on technological advancements and the basic functionality of the headsets. He said it was frustrating to hear from people inside Meta who found the Quest 2 headsets so unreliable that they refused to use them for work or demo them for people outside the company.”

Has anyone tried accounting for the tens of billions of dollars Facebook is spending on this? This pattern—massive outlays followed by poor, possibly-rushed possibly-underpowered workmanship—sounds remarkably like corruption.

> This pattern—massive outlays followed by poor, possibly-rushed possibly-underpowered workmanship—sounds remarkably like corruption.

I worked at a medium-sized company that went through a phase of hiring ex-FAANG people, thinking we'd improve our quality by implementing FAANG practices.

It did the opposite: The ex-FAANG people were absolutely masterful at self-promotion, office politics, and collecting wins for themselves while shifting blame for anything that didn't work.

The strange thing was that many of them were actually good programmers when it came down to it. It seemed like they had been conditioned by their FAANG employers to put self-promotion and survival above everything else, which turned into an extremely toxic trait once they were removed from FAANG managers who were playing the same game. The company had to steadily ratchet down the levels of trust and independence granted to teams, while steadily increasing the amount of management oversight and process to keep them within bounds.

Now whenever I see famous builders and founders leave FAANG companies out of frustration, I get it. The big tech company game has deviated very far from execution.

IT projects fuck up all the time. Poor execution is baked into the human DNA. It doesn't take malice or crime to fail to be excellent. Neither the Newton nor Google glasses failed on corruption. Carmack is smart but he also failed to make his rocketry mark. Maybe individual brilliance in code is less important to execution of things than we (he?) thinks? Or maybe he only picks targets with hard end-goals versus the easy marks?

Or the typical poor execution at extremely large tech companies. Where various people have incentives that aren't aligned with the overall objective of making a good product. Instead people are incentivized to empire build and be self serving.

> It pains me to hear people say that they don't even get their headset out to show off at the company because they know it's going to be a mess of charging and updating before they can make it do something cool.

He has a point. The tens of billions of dollars seem to be mostly dedicated to adding invasive crap that makes the experience worse.

After the last update I had to "sync my Meta account" before I could do anything. I had to log in on the PC (with my headset on) and enter a code shown on the headset on my PC. Yea, that's fumbly.

Oh and after the latest update may play area is apparently too small for room scale mode, it always forces me back to stationary mode. What a mess.

This was after using the device just fine for two years without such boondoggle.

All the money in the world can't buy talent. If the talented people don't want to work for or with you, then you end up just burning through cash while the untalented/less talented deliver a subpar product.

The more I understand about how many risks and problems are involved in running an organization as large and with as high a social profile as Meta, with all the accrued obligations and entrenched loyalties and overpromising managers that glom on over time, all while building new unproven technological infrastructure, the more sympathy I have for how blurry the line between corruption and implosion is.

I don't think anybody trusts Facebook at this point, and the idea of being forced into a pay for play virtual world controlled by a single company is understandably scary and deserving of criticism, but at the same time I've done social VR and I truly and honestly believe it's a vast improvement over any other existing kind of remote interaction. I actually get and believe in the vision, it just seems like the execution is really bloated/bad and they tried to way over abstract things/make things way too big way too early. I could see how spending that amount of money would be easy to do if you're trying to create universal abstractions for making, selling, creating, stitching services for an interactive set of 3D environments together that is easy for people to make "apps" for, building data centers, accounting for latency, building out supply chains, subsiding hardware, doing R&D on tracking, doing social research, and dealing with entrenched employees and middle managers accustomed to "the cushy tech worker lifestyle".

You really have to try social VR to understand how much better it is than video chatting or text channels. There's absolutely something to it. There's so much more information in the way people interact about what their intentions are, how they feel about what you're saying, what they're interested in, etc when you're interacting in VR with a 3D avatar.

All that being said, the virtual real estate and clear corporate "branding" angle is really gross, regardless of whether the money gets shady/you could really call that "corrupt". I prefer the wide range in quality but genuine effort and creativeness you get with a free and open ecosystem like VRChat.

I have a quest pro, and it’s pretty awesome. Zuckerberg made the argument that Microsoft spent about 10 billion on development of the first Xbox, and that turned out to be a good bet.

> sounds remarkably like corruption

In the sense of external entities illegally hoovering up cash, I doubt it. For me this seems more like Soviet central planning, where real-world success is secondary to conformance to big ideas imposed from top personalities. It can work when the idea is correct, useful, and resonant with the people implementing it. But when the idea's off, political realities prevent honest feedback from rising up to the level where it can have an effect, so everybody just goes through the motions and things slowly fall apart. A reality explained by the ancient bit of capitalist samizdat, "The Plan": https://web.mnstate.edu/alm/humor/ThePlan.htm

If I understand the situation correctly, there are different headsets aiming to solve different goals, and there are projects to integrate each headset into a headset that has all the desired features.

I think it's a push to be first to market with a headset that serves as the foundation for a monopoly on metaverse-like platforms.

In that talk he was talking about Horizon specifically and not Quest 2.

Like Tesla FSD, it's a knob for the company's share price. All VC-adjacent moonshots are. They're the encroachment of symbolism into capitalist systems of value.

John Carmack leaving your organization with a resignation letter like that is an undeniable sign that it is permeated with the rot of ineffective and self-deceptive middle management. If you've followed John's career you know that he has enough FU money and respect that he doesn't care about performance reviews, your promotion ladders or departmental politics. As he writes, he cares about shipping cutting edge technology nobody else thought possible that delivers exceptional value, and brand new experiences. His ethos is egos be damned, and breaking down organizational and cross-corporate boundaries.

Carmack should have been invaluable at Meta as the speaker of hard and uncomfortable truths. Likely Zuck & Boz have no idea whats really going on with their technology and are insulated by layers of self-deceptive status reporting. Once you step into management and away from the visceral struggle of building and using nascent technology daily, you are likely to make the stupidest and most illogical decisions. Setting that hard shipping deadline to coincide with your conference feels like you are making your team more efficient, but likely you are also forcing them to take on eventually crippling technical debt, or make rushed and ill-informed choices.

Its likely that top engineering talent, the ones who actually know how to build the next several generation of VR technology, already left or will leave soon. Those that remain will drown in so much clueless bullshit that they will pop out another VRML or Second Life and call it the metaverse.

It is a fad. No one really wants this. Just like 3d TVs the novelty fades away after a few hours.

I see a lot of negative comment on the other thread and on twitter so I feel the need to comment as well. It’d be great to get the perspective from someone who worked there, but I can offer mine as an outsider (worked on a different team at fb).

Back then, I would religiously read every workplace posts from Carmack (that wasn’t hidden to other orgs). I loved everything he was saying, I always found it really insightful. It also seemed clear (much before he went part time) that he was saying that on the outskirt of the project.

There’s a big culture of flatness at fb (and now meta). No titles are public, and you’re supposed to respect everybody’s opinion in the room. You’re supposed to lead without authority: by convincing people. You’re also quite free to explore things, as long as you can make a case for it.

There’s upsides and downsides to such environment. You can really multiply yourself if you create trust and clout. On the other hand, authority and a big title doesn’t always gives you room for directing a project.

I was always wondering how effective Carmack was going to be in such an environment. He doesn’t seem to be the type to lead, but I can see this happening in a small team, but an entire org that’s growing extremely quickly? For the kind of things he wanted to happen you’d have to make sure to hire people who cared about exactly the same kind of stuff, which doesn’t really happen when you’re in a diverse environment. Extremists must then spend their time pulling the group in one direction or the other.

I worked at Meta. Yea, there was a promoted culture of flatness, but that only applied to the IC side. Within engineering there are separate levels but everyone has the same title "Software engineer" so that (supposedly) level shouldn't be a factor in whose argument is correct. In practice, everyone knows who's high-vs-low level, so it's mostly a charade.

On the management side, there are clear titles, reporting hierarchies, and areas of ownership. They decide what gets prioritized, staffed, and shipped. As an IC, you can make technical decisions and recommend product ideas, but you'll never ship functionality to users without PM, design, and eng leadership approval.

Within RL, like everywhere else at Meta, there are VPs and Directors whose job was to set strategy and create roadmaps. RL had a multi-year roadmap aggregating roadmaps from all the sub-orgs within RL. There's no world where a single IC, even at Carmack's level, is going to subvert that hierarchy. The job of a senior IC is to influence and check their management counterparts to make sure they're prioritizing the right things. Clearly he wasn't able to do that to the extent he wanted.

When I was there a couple years ago, I looked into joining RL a couple times but got advice from coworkers that the rotating carousel of leadership led to changing roadmaps/priorities, so it was a thrashy place to be.

> You’re supposed to lead without authority

John seems to be particularly calling out that he was not able to influence the people with high enough rank to make decisions. So it sounds like he at least thought there was an authority structure there wasn't as flat as you describe.

I feel like Metas 'flatness' is revealing itself as suboptimal now that its stock is tanking. While the product is generating money and productivity doesn't really matter, the organization can afford to have masses of engineers doing not much of anything. But when things are going bad and clear top down organization is needed, then this culture suddenly turns into massive friction and might end up killing the company.

I don’t understand what the problem would be for him if this was the environment. He’s respected as a technologist and had strong opinions on technology. He’s not merely a tech. savant without social skills.

Again, this assumes that this was what the environment was like. Not that it was just like that in name only.

I’m a strong believer in dictatorship at companies. Every team has someone with ultimate decision making power, as does every project with a project leader, and above teams are leaders who have authority above them, and so on. Of course there are meetings and discussions but there is always someone with the authority to make a decision that everyone has to respect.

Honestly it benefits everyone. ICs like that authority is clear. Effective companies are mission-focused and they are not democracies. FB sounds extremely bloated and ineffective.

I've come to agree, 100%.

Every company needs a benevolent dictator with exceptional taste who is addicted to simplicity. This can be a platform architect, this can be a product person, etc.

But without that person, with design-by-consensus, you end up with a company that is addicted to complexity, building a product that looks like it was disparate components that have been continually stuck on the end with duct tape.

This is why we have the bystander effect, and why in EMS, you need to point to someone and assign them the responsibility of calling 911. There needs to be one person accountable, otherwise there is no one accountable.

Phrases like “benevolent dictator for life” have become pretty widespread, and they are cute, but I think the slightly obscure the issue or focus on the wrong thing.

A dictator can function with a small, powerful core of supporters (usually the military and then some key demographic or powerbrokers depending on the situation), and then keep the rest of the populace in their place through violence if necessary.

I think when we talk about somebody like Linus or Guido, they aren’t really dictators so much as consensus leaders. Everybody might not agree with their individual decisions, but there’s widespread agreement that somebody has to lead, and everyone can agree that there isn’t anyone who can do a better job.

And we should also note that Democracy isn’t some self-perpetuating system that can actually impose itself. Functionally, we don’t have a democracy because there’s 51% support for the leaders. We have a democracy because there’s broad consensus in the idea that, while we don’t always agree with the individual decisions of the democracy, we don’t have a better way of making decisions.

Honestly I don’t think it really matters so much how the decision making process works. Functional systems are ones in which the decision making process has broad consensus support. It could be tied to an idea, it could be tied to a person, it could even be tied to some abstract idea like following the rules of a religion. What matters is that people broadly think the system is good, follow it when nobody is checking, and are ok with following the spirit of the rules rather than getting hung up on process and technicalities. In open source, these tend to be personalized systems because somebody will just start a personal project and it will grow, but that’s very tied to the nature of programming and the fact that an individual can easily get a minimum viable product for many programming ideas.

I’m blanking on the name of the essay, but there’s a compelling argument for the only viable alternative to a Benevolent Dictator being a very thoroughly and precisely defined shared terminology of the underlying product/software system. Basically you can achieve a shared vision only by relentlessly checking the assumptions of the group and codifying everything, or using a singular visionary as the stand-in for a cohesive shared vision. Hoping someone here can share the original essay since it does a much better job than I can…

Agreed. Hierarchy is important. Someone asked me to do something? Directly above me in the hierarchy - done. Someone at my level or under me? Add it to the backlog along with 100s of other things.

The trade off is decision cycle time. In a dictatorship, the dictator (and getting their attention) becomes a bottleneck.

Sure you can have a good dictator, but there is much greater room for a bad one.

Carmack (by his own admission) doesn’t want to be a manager, he wants to be a programmer. So he would probably not make a good dictator. He might have had less sway precisely because someone above him is a dictator who didn’t like his input, no room for decent in a dictatorship.

Finding good devs is hard enough without a dictatorial style of leadership. I find that people like democracy, for some reason.

Like how twitter is run right now

Sounds like the perfect structure to end up with an utterly dysfunctional company ruled by the Peter Principle.

How does one effectively filter for these types of companies while doing a job search?

Basically Carmack is frustrated by:

1. Rate of progress on AR/VR at meta

2. People within Meta not "Giving a Dam!" / Poor decisions made within the org

3. Poor quality of the execution which I guess is related to #2

Relevant quote: "It pains me to hear people say that they don't even get their headset out to show off at the company because they know it's going to be a mess of charging and updating before they can make it do something cool," Carmack said at the time. "VR should be a delight to demo for your friends."

As an early and complete VR adopter I have ranted in the past on how fast VR loses it's luster. But for me the reason is not setup (charging, updating, putting everything into place, et cetera), it's software.

The good VR games are really good, games like HL Alyx. But after almost 10 years of VR usage I still can't name more than 3-7 pieces of VR Software (games included) I would put into the "worth it" category. Anything else is just very short lived, feel like tech demos. Fun? Sure, but ultimately worse than regular software/games. Gimmicky overhead, so to speak.

Hence why my VR setup is collecting dust most of the time.

Seeing Facebook's vision for VR, which is just rebuilding products we had as far back as 2003 (Second Life) and even 2014 (VR Chat), but worse and riddled with ads and darkpatterns, is what has made me lose even more of my interest. I have to add that I was an Oculus owner in the past, I am now an Index owner, but what Facebook does is still relevant because they are the "VR believers", pouring the most money into it.

In short: There is just nothing interesting happening, my VR setup is collecting dust. Demoing it to my friends is basically the only thing I do with it nowadays but they lost interest too. I don't agree that it's a hardware problem, for me it's a software problem. Also: too much "vision", too little actual determined (software) projects.

This was my gripe about all console games. Real gamers that are daily users have their systems kept up to date, but I'm not one of those. I'm one of those that plays a few times a month, and only when I have nothing else to do. Because of the infrequent use of the console, I'm guaranteed to have to do an update on every use. The update alone cuts into the time allotted for gaming. So I play even less and less to the point that after my last move (3 years ago), I never even plugged my console in.

So I totally sympathize with the loss of enthusiasm for the delays from forced updates.

I rarely put my VR headset on because I have to COMMIT to it and stop doing everything else; i.e. no host OS's GUI is VR-friendly.

Why did smartphones supersede computers in usage? and laptops supersede desktops? because of the low commitment barrier: something you can pick up and put down at any time without necessarily pausing other activities, is always going to be preferred by the masses.

So all current VR is sitting behind a user-commitment barrier, often collecting dust. Looking forward to see how Apple will tackle this problem.

The whole metaverse concept Facebook is doing honestly reminds me of Neom. Unaccountable god-king has a wild idea based on very real trends, has no real idea how to do it but a limitless supply of money from other things, and the whole concept gets taken over entirely by corruption

That’s the correct quote. The Quest Pro with its wireless charger actually largely solves this problem. Try not using an Android phone for a month, and then use it. It’s the same nightmare (after all the headsets are Android devices with internal batteries).

It’s an important gripe, but it will be solved. I think there are more fundamental issues. It’s hard to explain, but even with superphysical level of improvement I don’t see people enmass wearing VR or MR headsets the way we use phones, airpods, and to lesser extent watches.

I feel John's pain, after a decade of consulting and working for big tech companies.

I think though the major skill at a large organization is not speed of delivery, or even amazing software, but getting everyone bought in, and rowing in the same direction. That's really hard, and 90% of it is emotional labor. That's actually what a Principal Eng or Director does at a big company. Then if you can actually turn the 'Death Star cannon' of a large org at something, it's really powerful. But it can certainly be an exhausting skill.

I think folks who are great software builders, that thrive at small companies can sometimes fail to appreciate those skills. Conversely, 'big company' people look at the crazy, yet productive, ways of small companies and roll their eyes at their lack of "maturity". Both sides require very unique skillsets, and I'm appreciating, often disjoint sets of people that thrive in both settings.

John Carmack did a 5 hr podcast with Lex Fridman recently [1]. Moving away from Meta seemed like something that had been happening slowly over a long period of time.

[1] https://lexfridman.com/john-carmack/

Lex's guest list is absurd. I'm honestly not sure how he's managed to pull it off.

The real problem here is the memtaverse. This is a solution to a problem no one has and it's costing a vast amount of money for, well, nothing really.

Why does it exist? Simple. Meta needs a new monopoly since FB usage is declining and IG is getting eaten alive by Tiktok. Messaging (ie FB Messenger, IG Direct, WhatsApp) is not enough.

Meta has long seen VR then AR as the natural evolution from text -> image -> video. I think it's clear that the VR part of this at least is in error. Personaly I see VR as never being anything more than a niche. Somehow spending $20B+ a year on that without any kind of product-market fit or a vision for what problem this will solve for people is the problem.

AR is way more likely to have a future but the tech isn't there yet and there are doubts it'll be anything more than a niche either. This is a deep topic but projecting things onto real vision isn't exactly simple. Even something as simple as the color black is a problem. Focus is another giant problem.

For anyone surprised how a project can spend billions without producing anything, this is classic big company poorly defined project type stuff. A project will expand to fill available resources. Writing a blank check just increases the head count. It doesn't produce more just because you have more head count. People without clear direction will invent fake work for themselves. They'll solve non-problems, creates frameworks, add processes and so on as necessary.

Disclaimer: ex-Facebooker.

For those who don't want to view Facebook post (or can't login to it). I created pastebin of his resignation post: https://pastebin.com/Jk5TrGch

Also, the other discussion (700+ comments): https://news.ycombinator.com/item?id=34022484

Thanks - we merged the threads, and I included your URL in a pinned comment.

Normally I would have pinned your comment instead (sorry), but I needed to include the previously submitted URL and obviously I'm not going to edit your post to put that in.

Switching to reader mode also works.

His unscripted talks were amazing. Everytime I watched them I was thinking "ok, the official Meta announcements are corny, but there is still someone that still cares about what is the actual end result;" that's why I was still semi-positive about Meta.

Well, I guess not anymore.

I know the whole Oculus / Quest has a reputation for being a gimmick, but those talks were the ones that inspire me to get a quest 2 in the first place, and I love that device, It's far from perfect and requires a lot of 3rd party accessories to be comfortable, but it made me feel the same way as when I was a kid and play a new generation console like the PlayStation for the first time, making me say "wow, this it's the future".

Plus PCVR (something I feel meta lately doesn't push as harder as they should) with the quest it's just amazing, being wireless makes up for the loss in the visual quality of something like an HP Reverb or Vive.

I feel sad about John his departure, he was one of the voices who pushed sideloading for 3rd app developers and those game communities are amazing, creating games that have given me more entertainment than most of the horizon-verse apps.

Even in his last talk about the Quest Pro he was straightforward telling people "This device it's great, but still needs a lot of development and has some rough edges, like FOV rendering is not going to happen soon for most games" to have people don't over hype their expectations... while Mark and the official presentation was using pre-render avatars with legs for a feature that didn't exist.

He was wholey un-corporate. So refreshing and real.

Really shows what that brings to a corporate shitshow that is Meta.

Were they? I stopped watching after not being able to bear the stuttering graphics anymore.

https://archive.ph/sfnzN

We've since changed the URL from https://www.businessinsider.com/john-carmack-meta-consulting..., which that URL is a copy of. to the letter that everyone is responding to.

Nothing could be a stronger indictment of Mark Zuckerberg's ability to lead teams and create innovative products than the fact that he wasn't able to utilize John fucking Carmack properly.

Even with all the money in the world, Zuckerberg is just too conventional, uncreative, and lacking in genuine enthusiasm. He's been totally unable to create the kind of esprit de corps present in every great team.

He cloned and acquired his entire career and it's made him weak. He was never forced to actually get good.

What inspiration there is at Meta today was brought there by Palmer Luckey and John Carmack years ago. And, since it can't be cloned or bought by Zuckerberg, it has simply dwindled away year by year.

The upside is that Oculus launched the VR industery. And now it's just a matter of time until VR evolves from a toy to a powerful tool. Maybe Meta will eventually make it happen, they certainly have the money, but more likely Apple or a startup will give it The Big Push.

The skills and talents needed to lead and influence a big organization are different from those needed to influence a small team. Once the team size surpasses Dunbar's number, a phase shift is required into a qualitatively different way of leading, in order to be effective.

A few years ago I took a new role in my org, a team in a big tech company, where I was expected to influence the technical direction the key infrastructure that supports our mission. I had no direct authority to tell anyone what to do, except the small team that reported to me. But I had the endorsement of the bosses and a reputation as a respected technical voice from years as a senior IC before moving to management. No problem, I thought. We're small, and I know everyone and they know me. Those things had always been enough.

During that time, though, the team headcount grew fourfold, we had been less than a hundred people and now we're three hundred or more. We stopped calling ourselves a "team" and started to say "org" or "department". It was impossible to have personal relationships with everyone, and the ways I used to influence change stopped working. I became ineffective. I could influence individuals, but without defined processes and management systems that cemented my authority I couldn't influence the org efficiently anymore. I could see things going in bad directions, and I could get meetings with leaders and give them my opinion and recommendations. The would listen and nod in agreement, but the ship wouldn't turn.

The result for me was many months of near-burnout, the feeling of shouting into the void. What saved me was leaning on project and product management, and stepping back into a role of setting and influencing requirements and priorities, where I still have a voice people listen to. I use my one-on-one relationships to preview my vision for our direction and get feedback and buy-in from the other leaders, but the PMs manage the team-wide communication and execution. After almost a year of this, I think we're starting to be back in a good place where we have a roadmap and know what we're doing, but I also know we'll never be as nimble as we were when we were 50-75 people.

I actually thought facebook was doing a ok job with oculus until the whole 'metaverse' social push happend

it feels like the people invovled are serverly deteacted on what people want out of VR in general

it's hard for me to imagine someone like carmack prefering to work in a envinronment like that (tearing down corp/product walls) vs building his own team and trying to solve technical challenges

I just want cool hardware people can hack on. I couldn't care less about the metaverse worlds that they were building but I'm giddy as a child on Christmas at advances in the hardware lately.

> I actually thought facebook was doing a ok job with oculus until the whole 'metaverse' social push happend

Spending billions of dollars on a subpar MMO game when better ones cost tens of millions at most to make - is bad?

Forcing a mandatory Facebook account is "doing an ok job"...?

> I actually thought facebook was doing a ok job with oculus until the whole 'metaverse' social push happend

the only good that they were doing was decent hardware at a decent price...

Metaverse as a general concept is not a bad aim and Meta seemed to have some interesting ideas.

But their presentation was clumsy at best and they had miles to go to convince critics that they were not aiming for a very dystopian interpretation of the concept. They failed spectacularly on the PR side. Probably enough to make their own engineers lose hope regarding both ethics and executive management.

I think they might be right in the long run but that they are pushing it too hard when it’s not ready yet.

People probably do want to collaborate in virtual rooms at work, but only when it’s better than video calls. Which it currently isn’t. And they don’t want to do it in a world controlled by zucc

I'm going to link this because it's relevant. Safi Bahcall wrote of this phenomenon: https://mentalmodels4life.net/2021/01/04/safi-bahcalls-innov...

In very large organizations (>150 people) incentives shift dramatically because it becomes more valuable to engage in politicking than actually delivering value.

Carmack has perpetually worked in smaller orgs as an IC and has a reputation for being difficult (that is, a reputation for actually giving a shit).

Based on Bahcall's hypothesis, one potential solution would be to break the org into smaller units and create milestone based incentives like large team bonuses centered around performance bonuses

"He is a well known and regarded game designer, who moved to a new consulting role at Oculus in 2019"

Err, WAT? In the context of gamedev, game design is not what I'd say Carmack is known for. He's firmly in 'down in the weeds implementation detail and optimization' territory. How does one write anything about Carmack and get something so fundamentally wrong about who he is and what he's done professionally?

Game designer credit in this context (id software) generally goes to Romero, if anyone.

Edit: Just look at Carmack's wikipedia page FFS, he hasn't been credited for design since SOFTDISK: https://en.wikipedia.org/wiki/John_Carmack#Games

People take “game designer” to mean “anyone who worked on one of them thar computer game things”.

It's business insider so

That's like calling Schwarzenegger a GYM owner... I mean yeah it's not a lie, but it's underselling a lot.

There is a pay wall, so I haven't read the article. I just want to put it out there: Did this seem predictable to anyone? I had the sense he was unsatisfied and Meta's public perception, particularly in Carmack's department, seemed to be plummeting. It seemed very likely that this would happen, but maybe I'm projecting too much. I imagined myself being deeply ungratified in his role.

This blogpost from 2014 has been an almost perfect predictor of the last few years of the Meta disaster. It misses some stuff, but it's not really wrong anywhere beyond some minor details. It calls out Carmack as being the canary that will ultimately predict Oculus/Meta's future.

https://assayviaessay.blogspot.com/2014/03/virtual-spaces-re...

It always seemed like a poor fit to me. From the begining, Facebook has always been an unctuous, slimy organization and Carmack seems to be a relatively high-integrity human. I'm honestly surprised it lasted as long as it did.

His note to Meta: https://archive.vn/t6zQp

He hasn't seemed happy since the beginning. So this was a surprise to me. I thought he would leave after something vested, but that doesn't seem to have determined the timing.

The whole message posted to Facebook:

https://www.facebook.com/permalink.php?story_fbid=pfbid0iPix...

Wow can definitely relate to this being in a high level IC position, but not having enough influence on management to change the direction of the company. You sit on the sidelines knowing a decision is wrong, and after having said your piece you watch it play out over years, turning out exactly as you expected it to.

Bah humbug.

> [edit: I was being overly poetic here, as several people have missed the intention. As a systems optimization person, I care deeply about efficiency. When you work hard at optimization for most of your life, seeing something that is grossly inefficient hurts your soul. I was likening observing our organization's performance to seeing a tragically low number on a profiling tool.]

as someone who spends a lot of time in a profiler, this resonates with me. the irony that i work in JS/TS is not lost on me, but most React apps make me sad, most node_modules make me sad. V8/JSC/SpiderMonkey are amazing JITs, and seeing them get bogged down by inefficient JS is painful. i see many devs jumping to Web Workers or even WASM when in fact their existing JS and algos can be orders of magnitude faster with just a tiny bit of forethought.

> it is the more personal pain of seeing a 5% GPU utilization number in production

yikes, is it that low?

I think the 5% was a metaphor for how the org is barely able to utilize its resources.

Also wouldn't React JS/TS be an instance where you are fully utilizing resources? Not from a raw machine performance standpoint, but from a developer efficiency standpoint. Using nested for loops instead of a series of JS array functions is way more efficient but it is not even close to worth it. Machines are so powerful these days also that sacrificing readability/maintainability for performance doesn't really make sense to me.

I'm interested in learning more about how to optimize JavaScript programs. Do you know of any good examples or other resources in this area?

So what do you use? Vanilla JS?

I am not surprised.

He is not leaving Meta to focus on AGI and build a new company.

This is other way around, he is so tired and disgusted that he is throwing the towel and focusing on something else.

This is not a vote of confidence.

It is just impossible to innovate and deliver at mega corp like FB. That’s why acquisitions become the only mean through which they can “pretend” to grow.

Mega corp are good to rest and vest. If you wanna build stuff you should work at small innovative companies.

Meta has made the mistake of trying to build the killer app, rather than building the killer App Store.

Apple software is garbage for the most part; they outsource software dev to the entire universe and take a 30% cut.

Meta should be doing the same thing but instead they are spending 10B to get people to play horizon world, a wii graphics chat world app.

They should be incentivizing developers of games like world of Warcraft to build a quest 2 interface; 100M addicted users who want a more immersive world.

This isn't going to make me friends. John Carmack hasn't be relevant for years, and this doesn't move me one way or another. There was a time when game engines were simpler and the work of incredible engineers who could make something happen five years sooner, but we're past that, and I'm not sure Carmack is.

> it is the more personal pain of seeing a 5% GPU utilization number in production. I am offended by it.

For a high-level badge post, I'm not sure why he's calling out a low-level metric like this. It's also a complicated metric. Keeping GPU utilization low improves battery life. Delivering jaw-dropping graphics might have been the goal in 1993, but we're past that, and for all its flaws, I doubt Wii-level graphics are the reason no one uses Horizon Worlds.

Which brings me to why he left now. The company doesn't want Quest to be just a gaming headset any more, but Carmack does. It's a fair disagreement to have, and investors are even with Carmack on this, but a mature leader would disagree and commit or quietly step aside. Carmack isn't a mature leader, he's a talented, high-level IC you hire as a mascot.

Carmack has simply hit a ceiling that others won't let him pass. Carmack isn't in authority, and as such, he has to rely purely on influence. Hard to do when you are focused on technology, because technology and humans are quite different (yet also similar). He's making it clear that his political effectiveness is limited and he has no motivation to continue his fight. It happens. You only have so much energy to spend in any given time period, you have to spend it where it will do you and others the most good. Everyone eventually learns from these things. It takes time. And how each person defines success will be a bit different.

This was quite poetic IMO:

"When you work hard at optimization for most of your life, seeing something that is grossly inefficient hurts your soul." - John Carmack

https://www.facebook.com/100006735798590/posts/pfbid0iPixEvP...

Politics at work. The most infuriating part is when non technical people whose whole day to day work is about expanding their territory win through politics

Based on that hour-long talk, I'd say Carmack just had a fundamentally different set of priorities. He seemed to be focused on latency and scale, where Meta seems to be pivoting to fidelity and graphics.

He’s probably the best graphics programmer in the world so I would trust his read on the priorities over whatever Meta manager handles that project

Meta definitely isn't focused on fidelity and graphics. Their mobile VR push degraded VR graphics significantly.

> He seemed to be focused on latency and scale, where Meta seems to be pivoting to fidelity and graphics.

Which makes sense, you want your users to be able to really take in the graphical glory of the Metaverse.

From his complaints, I can deduce one of two things:

1. Carmack isn't a good leader. It takes a lot of "soft skills" to lead a large organization, and while Carmack has proven to be an unpaired engineer throughout his career, I haven't seen him build outstanding engineering organizations.

and/or

2. Carmack was fighting for influence with others at Meta VR, and couldn't get his goals take priority over others (Zuckerberg comes to mind).

I wasn't in reality labs, but all the posts from carmack over the years that I could see really really pointed at him experiencing both problems. Not only that, but people publicly both agreeing with his concerns, but also questioning why he was complaining about it instead of navigating the FB politics to lead the change.

He didn't have much of a role anyways. His title was "consulting CTO". He has primarily been running his own startup for a while now.

When a company like Facebook becomes successful it attracts a bunch of "execs" whose only job is to bleed the company dry. It's hard for people to do good work. That is why good engineers shouldn't join faang but do their own thing or join a startup which hasn't yet attracted the leeches.

People are saying “if John freaking Carmack can’t handle the politics, it must be bad.” And it probably is. But based on what little I know about Mr Carmack, I suspect his personality and incentives are unusually unsuited to navigating the politics of a massive bureaucracy.

I don’t really believe the line about Meta being the right place for it to happen. I think he very convincingly proved that it is indeed not so. Seems innovation does require some agile leadership, and not an entirely flat structure which feels a bit like steering a boat.

I think Carmack is a great engineer. Part of that is being great at communication.

And he is! Usually. But this paragraph is quite hard to follow:

> If I am trying to sway others, I would say that an org that has only known inefficiency is ill prepared for the inevitable competition and/or belt tightening, but really, it is the more personal pain of seeing a 5% GPU utilization number in production. I am offended by it.

I think what he means is that he is offended by inefficiency in itself. Not because of any secondary ills inefficiency causes but just because he is that way.

He must know that hist last post will be read by many. Even if it would not leak, and only circulate inside facebook many non-developers would read it. And even if you are a developer "seeing a 5% GPU utilization number in production" might mean nothing to you. I assume it is bad from the way he phrases it. Maybe it should be higher? Probably. But honestly who cares about the GPU utilisation if the app does what it should in a performant way?

He could have just wrote. "... pain of seeing an application waste resources in production." And then nobody has to spend mental cycles trying to guess what does 5% GPU utilisation means for him.

(And it is absolutely a guesswork. I'm myself working in a team responsible for the performance of a performance critical application and I love to see 5% gpu utilisation. My job is not to fill up the GPU, but to do something useful for the business. If an app can do that with only 5% of the GPU all the better!)

And then in the next paragraph in an edit he tried to explain himself. Probably because people complained that they don't understand what he is saying. As if you can alleviate confusion by explaining more. Instead of you know, fixing the source of confusion.

I'm a bit sad that he didn't had anyone help him edit such an important announcement before he posted it.

> Make better decisions and fill your products with “Give a Damn”!

Now that I love! I want that printed on a t-shirt. :)

Changing the source instead of adding an edit would make previous comments that were asking about the unclear part not make sense anymore.

The 5% figure is kinda shocking, that’s why he put a number. And it’s I think why people appreciate him, because he gets specific instead if writing a vague statement like “pain of seeing an application waste resources”.

I really don’t understand how it can be unclear that 5% is bad.

The thing that was actually unclear was people not understanding it was analogy for organizational effectiveness.

Yeah this was really odd to read?

5% gpu utilization on your battery hungry mobile device? That's not bad it's good/bad??

Low GPU utilisation is a serious issue.

Either you massively over-specced your hardware and should have chosen something cheaper and with less power consumption, or your graphics quality is far below what it could be.

Let's not beat around the bush: wasted resources are expensive, in some way or another.

He was talking about working on AI the other day. Artificial general intelligence "or die" type of challenge taking.

Guess he's getting full time busy with that.

As grateful as I am for his contribution to VR, I can't wait.

He's not only brilliant and no BS, but seems to have a way of picking out and focusing on highly relevant and fruitful threads of progress without losing any nuance, with whatever he dips into. It will be exciting to see what might emerge if he pours all his effort into AGI.

Also his talent seems wasted battling Facebook, if they don't want to listen to him, why bother.

Is there any reason Epic will not crush Meta’s attempt of VR?

They own unreal. They understand immersion. They want to have their own App Store eon mobile devices. They just don’t know “social”, but heck.. meta doesn’t anymore either.

Maybe GREE was right about seeing Facebook as their competitor after all.

Getting into social just gets you into a host of other problems. Why even bother. It's a high-wire act, and yes you get $$$ ad money, but it sure is vulnerable to any weird shocks.

I haven’t worked at Facebook scale but I’ve seen similar things at smaller companies.

A lot of dumb stuff gets greenlit because someone put a lot of attention and effort into making power point presentations, and no one wants to be the guy that says: this is a load of crap.

This is a strong signal of how the leetcode hiring process gives rise to a bad and inefficient work environment staffed with the wrong type of engineers, resulting in a highly experienced engineer to simply walk away.

I lost massive respect for Carmack after he testified in court that he downloaded ZeniMax emails after leaving for Oculus. He also claimed he rewrote all the code from scratch but it's possible much of the IP was in those emails. This is dishonorable behavior. https://arstechnica.com/gaming/2017/02/oculus-execs-liable-f...

"We have a ridiculous amount of people and resources, but we constantly self-sabotage and squander effort. There is no way to sugar coat this; I think out organization is operating at half the effectiveness that would make me happy. Some may scoff and contend we are doing just fine, but others will laugh and say "Half? Ha! I'm at quarter efficiency!"

This is true for any large organization, not necessarily Meta. Look at Google, Microsoft, Amazon and others.

I think he probably just watched his own keynote that he was forced to give through that Metaverse avatar. It was a much better experience as pure audio.

I'm very curious what he means about the software not being right. I haven't played with the Meta Quest but my experience with HoloLens 2 is absolutely shit on the software side. Microsoft used to be decent at software (this is opinion but I quite liked Win32 and MFC back in their day). The HoloLens 2 software stack is just ridiculous on so many levels (they try to shoehorn Windows and Universal Apps, for the actual innovative 3D apps, you just use Unity or Unreal???). I could not believe my eyes when Microsoft's developer documentation tells me to use a different commercial entities software framework for 3D app development.

This is again a controversial point (but hey ... so is the letter being discussed) .. isn't the software framework for the 3D rendering, spatial awareness the future software platform (what Windows was or Android and iOS are today)???

I'm curious if this is the issue Carmack seems pissed about. Clearly there is a story here.

This feels particularly relatable as my company has also been acquired. I now know how it feels to work at a large company for the first time.

I've always thought Big Techs where the best to build innovative products because they had resources. It's the opposite due to politics and bureaucracy.

Big Tech gets away because of existing products that were made when the company was nimble. Plus some products are capital intensive like Apple Watch or AI research.

The best organisational setup to make great products are startups. Politics tend to be minimal in early days. If the company is rapidly growing, there is no time for politics.

I think it was Erik Schmidt who said that when a company stops to grow, politics and bureaucracy settles in.

That's probably what's happening at Big Tech.

Ad-powered businesses can project growth by simply adding more ad placements in their products.

This reminds me that no matter who you are, solving problems in big companies requires an insane amount of persuasion work. Storytelling and aligning people on why something is essential becomes the job rather than the vision to solve it.

Corollary: if people don't agree on the problem to solve and its importance, solutions heavily tend to fail.

I'm surprised he's able to be this open about it. Wouldn't his contract have a non-disparagement clause?

Whatsapp's developer had a similar frustration. Why are they selling to big companies and expect to have complete control? They aren't being realistic. Big companies are going to do what they feel is best for them not the the other way around. There's nothing new about that. I guess it's too hard to reject the big payday.

Relevant: The Gervais Principle, Or The Office According to "The Office" https://www.ribbonfarm.com/2009/10/07/the-gervais-principle-...

From his tweets..

>I see no reason to expect that Meta could build a better CPU/GPU than Qualcomm (at much lower volumes!), and there isn't as much opportunity for full custom special units in VR as people think. This leads people to propose lots of MR features, which are of unproven value. https://twitter.com/ID_AA_Carmack/status/1603936133632905216...

I wonder if John is just not a believer in MR (mixed reality) that this is what actually led to him leaving. I could see him advocating strongly for pure VR, as its his bread-and-butter and where he probably believes is the most innovation. MR presents challenges and ideas he might not be well suited for.

You can find John Carmack's full post here: https://www.facebook.com/permalink.php?story_fbid=pfbid0iPix...

Some peoples faces are different when they are alone. Some people struggle with themselves just to present a normal looking persona in their everyday lives. For them it is good enough to live as stress free as possible. Their true underlying goal is not to go out of balance. Let them be.

I'm going to guess that there was an earnout that expired.

This is kind of a weirdly dramatic departure, as this was a union and a product that never made any sense on a napkin or the back of an envelope. Sigh. Hope to see him back on track and doing some cool stuff in the future.

Mark and Meta have been given a once in a lifetime chance to right their way by refocusing and ironing out privacy and feature set while Twitter burns its house down. I wouldn't squander it on the metaverse.

From his Facebook account:

I resigned from my position as an executive consultant for VR with Meta. My internal post to the company got leaked to the press, but that just results in them picking a few choice bits out of it. Here is the full post, just as the internal employees saw it:

------------- This is the end of my decade in VR.

I have mixed feelings.

Quest 2 is almost exactly what I wanted to see from the beginning – mobile hardware, inside out tracking, optional PC streaming, 4k (ish) screen, cost effective. Despite all the complaints I have about our software, millions of people are still getting value out of it. We have a good product. It is successful, and successful products make the world a better place. It all could have happened a bit faster and been going better if different decisions had been made, but we built something pretty close to The Right Thing.

The issue is our efficiency.

Some will ask why I care how the progress is happening, as long as it is happening?

If I am trying to sway others, I would say that an org that has only known inefficiency is ill prepared for the inevitable competition and/or belt tightening, but really, it is the more personal pain of seeing a 5% GPU utilization number in production. I am offended by it.

[edit: I was being overly poetic here, as several people have missed the intention. As a systems optimization person, I care deeply about efficiency. When you work hard at optimization for most of your life, seeing something that is grossly inefficient hurts your soul. I was likening observing our organization's performance to seeing a tragically low number on a profiling tool.]

We have a ridiculous amount of people and resources, but we constantly self-sabotage and squander effort. There is no way to sugar coat this; I think our organization is operating at half the effectiveness that would make me happy. Some may scoff and contend we are doing just fine, but others will laugh and say “Half? Ha! I’m at quarter efficiency!”

It has been a struggle for me. I have a voice at the highest levels here, so it feels like I should be able to move things, but I’m evidently not persuasive enough. A good fraction of the things I complain about eventually turn my way after a year or two passes and evidence piles up, but I have never been able to kill stupid things before they cause damage, or set a direction and have a team actually stick to it. I think my influence at the margins has been positive, but it has never been a prime mover.

This was admittedly self-inflicted – I could have moved to Menlo Park after the Oculus acquisition and tried to wage battles with generations of leadership, but I was busy programming, and I assumed I would hate it, be bad at it, and probably lose anyway.

Enough complaining. I wearied of the fight and have my own startup to run, but the fight is still winnable! VR can bring value to most of the people in the world, and no company is better positioned to do it than Meta. Maybe it actually is possible to get there by just plowing ahead with current practices, but there is plenty of room for improvement.

Make better decisions and fill your products with “Give a Damn”!

"VR can bring value to most of the people in the world, and no one is better positioned to do it than Meta."

I'm now concerned for Carmacks sanity...

Musk on VR https://youtu.be/Ai-gJQ99ci0

I think he has some good points. I don’t think the majority of people want something strapped to their faces. And for those who do, who can play longer than 1 hour without getting motion sickness?

VR is a fad that comes and goes.

Carmack addressing the leak of his internal memo, and notably, sharing his next goal of achieving AGI at Keen Technologies.

https://twitter.com/id_aa_carmack/status/1603931905539325955...

I have to login to facebook to view this?

No. It works fine without cookies.

"I have never been able to kill stupid things before they cause damage". Every developer who had to postpone needed refactoring to spend time building someone's pet feature that eventually got <1% use can relate.

even the best of us vaguepost, huh?

i can hope that if you’re in the company you know exactly what he’s alluding to throughout this post, but here on the outside i’d bet more that this is the type of letter where everyone reading it interprets his words in different, and often opposing, ways.

The metaverse is a castle in the sky nobody can make it what Meta or John Carmack wants it to be.

The thing about 5% gpu utilization makes me wonder if all this talk about training LLMs on multi A100 clusters is largely unecessary and that it might be possible for everyday GPUs to reach say A100 levels simply by going from 5% utilization to 90%.

He's not literally talking about 5% GPU utilization. It's a metaphor for the eng org being inefficient.

Honestly, I suspect that many people at Meta know that the whole VR/metaverse thing isn't going to pan out, leading people in that department to (say) spend time engaging in office politics rather than working on developing the product.

Two things made me raise my eyebrows about Carmack: the way he let Steve Jobs talk to him (even in his own memories, so perhaps it was even worse) and him joining Facebook in the first place.

I guess humility is even more virtuous than they say.

Carmack is a developer god, but against stupidity, the very gods themselves contend in vain.

The way some users talk about him on this thread sounds almost cultish and creepy. I'm not American and don't play video games, but the way he's talked about you'd think he's Alan Turing or something.

I wonder how much of this is the result of a over bloated product management arm.....

Does anyone else on iOS get a browser alert “cannot open the page, URL is invalid” when the page seems to load perfectly? Any idea what’s up with that? Not something I’ve seen from safari.

Isn't the core problem VR. It will never fly. I remember I had some glasses that you put your phone in. Worked perfectly. But VR is even more annoying than home 3D movies.

> I have a voice at the highest levels here, so it feels like I should be able to move things

John discovers that steering a supertanker is different from driving a motorcycle.

What is this new startup he will be working in? Has he disclosed?

https://techcrunch.com/2022/08/19/john-carmack-agi-keen-rais...

Could it be a communication or persuasion skill lacking? Usually very smart engs cannot convey their point to less Talented ppl that usually makes the decisions

The actual memo is at https://archive.vn/Lmu9g which is more interesting than the summary.

I'm surprised/sad he didnt decide to join Valve. VR is a big bet for them as well and I'd assume he'd have more freedom/power there than in Meta.

At his age and level, you lose all interest in having a boss.

Given that Facebook accounts are now to suddenly be converted to Meta accounts, I am almost inclined to assume that it has something sinister to do with this.

I'm happy for him. The AI field is making things happen. VR has potential, but as of 2022, almost everything about it is hot garbage.

Updated archive: https://archive.vn/TQg2r

Too much brain for so bad purpose. I am glad that Carmack makes his move to go forward. He is a brilliant programmer.

It took longer than I thought. The best timing for him was when Facebook purchased Oculus.

Zeitenwende.

Should never have joined the dumpster fire that is facebook. Dispicable company.

Can someone please paste the letter? I haven’t had fb account for years now.

Check https://pastebin.com/Jk5TrGch

I hope he joins Valve instead. Facebook was the wrong choice from the get go.

He didn't chose Facebook. At the time of the Oculus acquisition, they must have made him an offer he couldn't refuse.

With Carmack's amazing talent and vision, unless FB changed somehow, FB is no place for the great Carmack!

Keep programming brother, leave the BS politics to the plebs, and go back out there with YOUR team an fing kill it!!!!

Why Carmack was involved with Facebook/Meta when he joined was pretty confusing to me. Even if VR is pretty interesting technology to work on, where’s the joy in it if you’re effectively working for the Empire?

Keep in mind he joined Oculus BEFORE the purchase, he didn't chose Meta specifically, he just chose to stick around because he believed in the VR mission enough to try and make it work.

I don't use facebook. Is this reposted somewhere?

Full memo is reproduced here https://archive.vn/Lmu9g

Godspeed John! I sure hope your next project has legs! ;)

Can someone explain me the 5% GPU utilization issue

As someone else has mentioned, I believe it's a metaphor for Meta as a company. It's probably a little extreme at 5% but he's suggesting they're only working at that efficiency level.

For the monster hardware it has, Oculus (at least in games) can only rival a PS2 in terms of polygon count. Pair that with the 5% GPU utilization metric and you get that this device can handle a whole lot more load and it's just being used inefficiently. I'd even use the word 'wrong', people within Meta made 'good enough' libraries and called it a day, and he's mad about that.

This thing has billions of dollars poured into it, and all you get is PS2/PSP level graphics? No way, man.

Carmack of all people knows the kind of feats this machine is capable of, having worked complex optimization problems in the RAGE Engine and the old Quake engine too.

Good. He deserves better.

I still wait for when he moves to Valve.

The real world welcomes you back, John!

Bye bye. In to bigger and better.

Was always a bit of an odd fit

Read Carmack's full memo:

This is the end of my decade in VR. I have mixed feelings.

The issue is our efficiency.

Some will ask why I care how the progress is happening, as long as it is happening?

We have a ridiculous amount of people and resources, but we constantly self-sabotage and squander effort. There is no way to sugar coat this; I think out organization is operating at half the effectiveness that would make me happy. Some may scoff and contend we are doing just fine, but others will laugh and say "Half? Ha! I'm at quarter efficiency!"

It has been a struggle for me. I have a voice at the highest levels here, so it feels like I should be able to move things, but I'm evidently ot persuasive enough. A good Fraction of the things I complain about eventually turn my way after a year or two passes and evidence piles up, but I have never been able to kill stupid things before they cause damage, or set a direction and have a team actually stick to it. I think my influence at the margins has been positive, but it has never been a prime mover.

Enough complaining. I wearied of the fight and have my own startup to run, but the fight is still winnable! VR can bring value to most of the people in the world, and no company is better positioned to do it than Meta. Maybe it is actually possible to get there by just plowing ahead with current practices, but there is plenty of room for improvement.

Make better decisions and fill your products with "Give a Damn!"

For those without FB/Reader mode:

------------- This is the end of my decade in VR.

I have mixed feelings.

The issue is our efficiency.

Some will ask why I care how the progress is happening, as long as it is happening?

Make better decisions and fill your products with “Give a Damn”!

OG Facebook Post by Carmack:

``` I resigned from my position as an executive consultant for VR with Meta. My internal post to the company got leaked to the press, but that just results in them picking a few choice bits out of it. Here is the full post, just as the internal employees saw it: ------------- This is the end of my decade in VR. I have mixed feelings. Quest 2 is almost exactly what I wanted to see from the beginning – mobile hardware, inside out tracking, optional PC streaming, 4k (ish) screen, cost effective. Despite all the complaints I have about our software, millions of people are still getting value out of it. We have a good product. It is successful, and successful products make the world a better place. It all could have happened a bit faster and been going better if different decisions had been made, but we built something pretty close to The Right Thing. The issue is our efficiency. Some will ask why I care how the progress is happening, as long as it is happening? If I am trying to sway others, I would say that an org that has only known inefficiency is ill prepared for the inevitable competition and/or belt tightening, but really, it is the more personal pain of seeing a 5% GPU utilization number in production. I am offended by it. [edit: I was being overly poetic here, as several people have missed the intention. As a systems optimization person, I care deeply about efficiency. When you work hard at optimization for most of your life, seeing something that is grossly inefficient hurts your soul. I was likening observing our organization's performance to seeing a tragically low number on a profiling tool.] We have a ridiculous amount of people and resources, but we constantly self-sabotage and squander effort. There is no way to sugar coat this; I think our organization is operating at half the effectiveness that would make me happy. Some may scoff and contend we are doing just fine, but others will laugh and say “Half? Ha! I’m at quarter efficiency!” It has been a struggle for me. I have a voice at the highest levels here, so it feels like I should be able to move things, but I’m evidently not persuasive enough. A good fraction of the things I complain about eventually turn my way after a year or two passes and evidence piles up, but I have never been able to kill stupid things before they cause damage, or set a direction and have a team actually stick to it. I think my influence at the margins has been positive, but it has never been a prime mover. This was admittedly self-inflicted – I could have moved to Menlo Park after the Oculus acquisition and tried to wage battles with generations of leadership, but I was busy programming, and I assumed I would hate it, be bad at it, and probably lose anyway. Enough complaining. I wearied of the fight and have my own startup to run, but the fight is still winnable! VR can bring value to most of the people in the world, and no company is better positioned to do it than Meta. Maybe it actually is possible to get there by just plowing ahead with current practices, but there is plenty of room for improvement. Make better decisions and fill your products with “Give a Damn”! ```

TLDR: It’s everyone’s fault but mine

All we're seeing here is the continuation of people with the means distancing themselves from the toxic brand.

This and the excuses people make for still working there draw hilarious parallels with the Nazis.

Seems a little off to be calling out perf metrics when the whole system hasn't found product-market fit yet

5% is a metaphor. Read that part again.

Reply on news.ycombinator.com