iamstupidsimple

> £4800/yr is a shit ton of money? Things must be pretty rough over there!

For the average person in the UK, it definitely is.

In the north of the UK, dinner=lunch and tea=dinner.

> However, tools, you can make a killing selling simple tools with sexy UX on Mac.

I've been thinking about doing something like this recently but I'm not a heavy mac user these days. Any tips on what people are looking for in small tools?

Somewhat relevant: Chet Faliszek (ex-Valve) is currently posting videos about how the flat hierarchy works:

https://www.tiktok.com/@chetfaliszek/video/72814172402960335...

https://www.tiktok.com/@chetfaliszek/video/72814409728508593...

https://www.tiktok.com/@chetfaliszek/video/72814159128240653...

These missing TotW's are usually about internal tools that aren't open source.

And having execs who design the organizational structure around those goals is part of what makes good engineering :)

Crab bucket mentality. Those salaries are fine, what went wrong is Meta overhired. FAANG companies don't have enough work to go around as it is.

> 1) How can you write this whole article without saying "Kythe"?

Well, it does discuss Grok (the closed source predecessor) extensively.

@dang doesn't work, you're best off emailing hn@ycombinator.com

Just personal opinion, but if this product is as valuable as you're suggesting, then definitely try selling at a higher price point! - Your target market can afford it.

Valve is also a games developer, and they never pulled any of their old or failed games. You can still play Artifact or Ricochet multiplayer (if you really want).

Those are 'third places'. The workplace was always the second place.

Sibling comment has a good point, this number includes everyone from software engineers to kitchen staff. Very few of my direct peers have left and the Google is growing still.

I joined Google ~6 months before COVID. More than 60% the company was hired after my start date, so that 1/3 stat is probably higher.

A more generalised approach: structure emails in importance order, top down. So, actionable stuff at top, rationales further down, in descending importance etc...

This becomes more critical when talking to execs.

Shackling accountability is no way to run a team.

That being said, there's value in ICs contributing to a roadmap, if they're trusted. Being a 'leaf node' and closer to customers can make people more in touch with problems and solutions a business could implement (at least, that's the theory). But in most organisations the customer and business problems are well-removed from engineers / designers leading to lower quality insights from those people; but in either scenario it always seems like management is out-of-touch.

I as an IC would personally be fine having a bit of 'skin in the game' if the risk:reward ratio was more favourable, but that does seem rare at modern corporates.

I think it's a generational thing.

As a young person, work is basically the default way to meet people after moving to a large city. I have real friends outside work now but they all started as colleagues or friends of colleagues. I'd also prefer to work with people I can trust over anonymous colleagues any day.

I'm not saying it's the only way (it's probably not good or healthy) but this sentiment is super common among my peers. When we're older I suspect that will change.

Unfortunately that problem space is so large (every new line of fibre has how many stakeholders?) that a single individual or organisation might not be able to have any impact here. Which I think is the root of this question.

Even with high ping, application layer is probably the wrong place to solve this problem. We'll likely get email working as one of the first problems and be back to correspondence chess and the like. Even Civilization 5/6 works over email.

> It's gonna simultaneously trigger and require the economic destruction of a huge amount of office real estate investments to enable the change to remote work to become the majority.

Baring in mind this could be worse than 2008. It's not going to be pretty and workers will ultimately feel the brunt.

I believe it's part of (or at least funded by) Twitter: https://en.wikipedia.org/wiki/Bluesky_(protocol)

Wow, I was just thinking about this scene as well! It really is timeless. Especially the bit about measuring success by activity and not results.

Software engineering is one of the few middle class jobs available these days. Understandably it's now packed with people desperately trying to live a specific lifestyle, buy a house, etc... rather than a passion for tech for tech's sake. (Not placing blame.)

That fact along with the average piece of consumer software having such low perceived value (try getting people to spend 99p on the app store) and consumer software has become this awful race to the bottom of lock-in and advertising. I suspect business software is much healthier for this reason.

Economic pressures are why we've arrived here.

Figure out the core data structures and do this at different level of abstraction. I'm currently reading through bits of the TensorFlow codebase, and starting with TF's GraphDef[0] was a good place to start. This is great because it's high-level enough that I can understand how TF models are structured on-disk and I can follow the deeper threads (e.g. how does device acceleration work) if needed.

https://github.com/tensorflow/tensorflow/blob/master/tensorf...

The vast majority of computer science lectures from 20 years ago are still relevant. A ton has changed in terms of frameworks but the fundamentals are mostly the same.

> Can you give me an example of a successful long-term nuclear storage facility that exists today?

Sure. The earth is full of radioactive material (also known as rocks) that have been 'stored' successfully for millions of years.

Advertisers measure outcomes, not raw impressions. Because everything is auction based there's a natural market correction based on this, so impressions could e.g. be worth 1/3 less come bidding time.

It's called a "house ad" and is generally free for the publisher. They appear when there's no other inventory (like insurance ads). Usually double-click stuff is all negotiated by publishers and has little to do with AdSense or bidding.

> Telling the govt "too bad, you can't stop math" will backfire. The law can be used to force tech companies to literally stop doing math at scale.

There is no practical scenario where those who want to use e2e will not have that capability. Even if technology companies are totally banned from producing it domestically, it's trivial for foreign companies to provide the e2e software and supply it over the internet.

It worked fine in the past when the mainstream platforms were also free speech platforms. Scale matters and those websites are only terrible because they're bastions for everybody else who also got banned.