John Carmack issues some words of warning for Meta and its metaverse plans | Hacker News Reader

John Carmack issues some words of warning for Meta and its metaverse plans

arstechnica.com

54 pointsAlexandrB21 comments

I for one am excited to be able to visit Hell on Earth within my lifetime. This is usually a privilege reserved for the dead.

Carmack has put over a decade of effort into this, as a labor of love and a business, and he's a legendary hacker. I'd gander his ASM chops and s/w arch comprehension dwarf Zuck's PHP skills.

If he says it is hard, listen to his reasons. I'm pretty sure he knows this domain.

Is Zuckerberg even any good at PHP?

https://gist.github.com/nikcub/3833406

Was PHP any good back then?

PHP, surprisingly, has legs. I started using it at my job in circa 2000, then moved to Rails, then to Node, then back to PHP for another project. It had lots of problems in 2000. But it has been fixed over the years.

It's not the dinosaur people make it out to be. The main concern was the number of threads it creates, and how that limits its scalability. But modern *Nix cores can handle more than 1024 processes. Node definitely improves efficiency by using async architecture which reduces OS load, but PHP isn't the devil as some claim.

I think there's a treadmill habit of JS frameworks that make anything >2 years old seem out-of-date, but PHP has it's benefits... assuming one adheres to separation of concerns (something that was casually disregarded by the same pedants when they discovered React, Angular, and Vue).

Wow, that was a soapbox rant.

I understand how much PHP improved, hence I tried to imply past.

Here's his actual keynote speech:

https://www.youtube.com/watch?v=BnSUk0je6oo

As always, John is a master of communication and clear thought.

Also, Carmack's notes: https://www.dropbox.com/s/ruv739914jzquue/connect2021.txt?dl...

Adding “architecture astronaut “ to my dictionary, such a nice term! Unfortunately, Carmack’s explanation of why this is a concern for Meta is unclear to me.

I think the origin of the term (or at least the first time I saw it) may be this 2001 article from Joel Spolsky, https://www.joelonsoftware.com/2001/04/21/dont-let-architect...

I assume he is afraid of people with lofty high level ideas in charge of pushing out products for the metaverse that are just not fun or even usable because they started top down, missing end user detail, while Carmack would start implementing the details instead and maybe the high level stuff emerges, maybe not, but at least you have products users want to use.

I believe it’s meant to differentiate big ideas(hand waving) versus the difficulty in implementing the minutia of those ideas.

The big idea is important, but it’s the implementation which determines if something is successful.

The nomad and the iPod are both the same big idea - but implementation is different.

That’s what confuses me. What does he expect from Mark, understanding the technical details of VR and such? Or not trusting Carmack to do his job well in getting the details right?

He does oscillate between what he believes the company is doing well and what they are not doing well in his view.

I suppose the motivation here is that he does want the metaverse to happen, but he understands that if the company flubb it then it ruins it not just for their company but in the perception of the public’s eye - history is filled with examples of excellent ideas and technology that didn’t take off because they failed to engage the public’s imagination.

I guess this is his way of saying you can’t achieve this by halves.

It’s about how you can get stuck on thinking about and implementing great architectures instead of actual usable products. Like never actually writing code because you keep getting stuck in debates about what programming language to use (or even about creating your own language to solve all problems).

It seems to me that an awful lot of money that could be spent on cinder block walls, razor wire, and private security is instead fueling pipe dreams of avoiding those things with some sort of matrix-type escapism.

I would bet that they get more long-term traction out of the walls, razor wire, and hired guns.

The most telling thing about these projects to me are how sheltered the bubbles of existence are for tech execs and their financiers.

Do they think there won't be heroin or meth addicts because of VR? Do they think that neurotic kids raised on legal meth (ADD pills) and legal heroin (anti-depressants) will suddenly be happy to go to work and shuffle the spreadsheets because there's VR goggles in the middle? Do they think VR headsets will make people want the soylent green when there's not enough farmable land left to feed the world's growing population? Do they think it'll be harder for a burglar to knock them over the head and take their stuff when they're zonked out on sleeping pills and alcohol, being aimlessly driven around town by their hopelessly lost Tesla which just had its GPS antenna fail?

All of those things aren't abstract things that "just happen." They are created by the current culture. How ignorant must the current culture's tech execs and finance execs be of their own shadows, to think they can just put sunglasses on and make those things disappear?

Crime is mostly a policy not a technical problem.

Not sure what made you think anyone is trying to solve the crime issues with VR. There is nothing in the OP pointing to that.

That being said crime moves online too because that's where people are. The burglar may not knock your physical door anymore, it will knock some of your accounts.

Well, no 5G for you.

So, less "must be 3d" and "must be domain models", and more "make a series of killer apps solving interaction challenges, incrementally improving component technology". And yes it's meh that many 'killer' apps being designed for the metaverse are intended for audience-capture.

Reply on news.ycombinator.com