onion2k

I wish the page used the CSS :visited selector and provided a link on the index page, so we could easily see which one we've visited before or not.

That only lasts as long as your browser history. For example, Chrome only holds your history locally for 90 days.

Invading Australia is looking pretty tempting right now.

I don't think that's true if the reason a company has left the vendor for given model by making it hard to buy. Enterprise IT is enough of a pain in the butt that people will forego the new shiny to avoid the old painful unless it's genuinely better. As you say though, the best frontier model flips regularly, so companies won't go through the hassle of deploying a model if it's proved horrible to do in the past. They'll just skip that model because their current one is fine.

If there's a human who could write SQlite in an hour we should be paying them at least that much.

The fact that the swarm needs high quality docs to work from will keep most of the tech industry safe.

Working with someone who thinks Team Topologies is an instruction manual is not fun.

Absolutely XXXXXXX XX XXXXX XXXX because I XXX XX XXXXXXX XX XX XXXXXX even if the state XXXX XXX XXXXXXXXX XXX XX XXXXX XXXXXXXX.

(This joke would be funnier if HN allowed unicode Full Block characters)

Projects can be broken down into parallel pieces of work, so you should always be able to pick up another part of the project rather than switch to a different project if you're blocked. As the other poster says, the goal is to not be blocked though. The leadership team need to support the delivery team to get the work done as a focused flow of items with as little blocking and waiting time as possible. If you're getting blocked, that's a signal that the org is failing.

Admittedly this is a little idealistic, but if it's not the goal then the reality is always far worse.

I'm pretty certain that the one factor above all others that makes software delivery slow down is doing too many things at once. You should aim to be working on one project at a time in your team.

Taking on another project, or workstream, or idea, or investigation in a team is pretty much entirely downside.

- more context switching, which means people get cranky faster

- lower bus factor for any piece of work, until its 1 dev per thing and that things stops of the dev isn't there

- lower people count on work means everything takes how long 1 person can do it (2 people is twice as fast for parallizable work)

- slower dependency resolution until something as basic as getting a PR reviewed is a big deal

- fewer people seeing the problem means less experience brought to bear, which reduces speed gains from having seen this stuff before

And more besides. For every project the goal should be maximum parallelism.

A good webkit-based browser on Windows might mean people actually start bothering to test things in that browser family.

It can be done today but it's a little bit of a pain and it lags behind Safari by some margin. Run `npx playwright wk <url>` if you have Playwright set up.

The USA and Russia used pencils in the early space missions. They stopped because broken pencil points would mean they'd have graphite floating around, which is electrically conductive, and got sucked into the ventilation systems resulting in electrical shorts and fires.

After the Apollo 1 fire Fisher invented the space pen and sold it to both the USA and Russia.

This is a top-down incentive problem. If teams are rewarded for making things work, fixing problems, and keeping the lights on they'll pick up ownership of orphaned projects.

If they're only rewarded for feature delivery then they won't, and they'll push back on being given any non-feature work like maintaining things.

If AI replaces my job specifically, get a different job.

If AI replaces all jobs, I'll mostly wonder how rich people will survive without being able to make billions selling things to poor people, and then realize that capitalism isn't going to stop so jobs will be necessary. Everyone having a jobs is literally how society funnels money around. Jobs aren't going away.

Parsley is used by people who don't cook often as a garnish to make a meal look more fancy. The sort of thing you might do if you have a date coming over and want to impress...

It's got nothing to do with what most people actually do when they're working..

AI companies claim their products are generalists though, and that they can do a good job on anything you give them, so you can't say what people will be doing with it. "Generate an SVG of an bird on a bicycle" is a corner case certainly but if a candidate interviewing for a role claims they can handle the corner cases then it's totally fair to assess them on that.

Besides, if you move up one layer to "how good is AI at generating valid SVG markup of non-obvious things", pelican on a bike is actually a good test.

Richard Campbell did a great talk at NDC a month about about this - https://www.youtube.com/watch?v=eo7MEPgWGic

How to take your product to market has been the downfall of many, many startups. Building something is the easy bit - getting people to pay attention to it is far harder, especially if you're not someone who enjoys marketing.

Hit up your network and find someone who genuinely loves marketing a new product, and ask for help.

However... Also note that the startup market has changed a lot over the past couple of years. AI has made people shift from "how can I pay to fix my pain point?" to "How can I vibe code an app to fix my pain point?" Marketing in those conditions is much, much harder than it has been in the past.

It's so tempting to drop the text of that paper into Pangram. :)

There will come a point when the price hits high enough to justify the cost but that also means higher costs to the end user.

This makes me wonder if it'd be worthwhile starting a company to capture it now, and just stockpile it until it's rare enough to be able to use my stockpile to control the price. The DeBeers diamonds playbook applied to helium, or maybe the Peter Thiel build-a-monopoly-to-win approach.

If you switch on the 'Supporting Evidence' on that site, it seems to be basing it's opinion on three things:

- Use a descriptive triad of "reviewing, directing, and course" (it incorrectly misunderstood 'course correcting'). That's not common in writing but humans do do it occasionally.

- Using the word 'thoughtful'. I don't understand that as evidence of AI.

- Using the words 'Book Apart' together, which would be a clear AI signal if it wasn't the name of a publisher of short books, and being used in that context in the article.

I don't think you should put much stock in the output of pangram.com.

The fact that Open Systems hold a trademark on "Open Systems" is less a signal that OpenAI should be allowed to hold "Open AI", and more a sign that Open Systems should start considering changing their name because they'd lose their trademark if someone disputed it.

The author should probably read The Checklist Manifesto. It's a great book that has some excellent advice about getting things done with higher quality by, not surprisingly given the title, using checklists.

Vibe-coded apps probably have loads, but mostly because they're using less capable models than the people who're doing the bug-hunting. Once vibe-coders are using models like Mythos too you should expect the number of bugs in vibe-coded apps to collapse quickly, because the LLM will write the bugs but will also fix them (assuming the system prompt tells it to.)

I'm willing to bet that "Daily Active Users" is one of LinkedIn's H1 OKRs. Putting a daily game on the site earned an SVP a fat bonus.

The main thing Claude knows about me is that I'm incredibly bad at my job and have to ask for help a lot. If you were to talk with my colleagues they'd tell you this is not a secret.

These people are already massively rich and likely have opportunities to join any tech org that's growing, and any of those opportunities could make them even richer. It doesn't really explain why Anthropic is their apparent best option.

It's probably just the most interesting and challenging tech company right now (that's hiring high profile roles?). There's lots of other companies doing interesting things but if your expertise is software applied to a complex domain, Anthropic probably comes out on top.

I've also seen an increase in merged PRs, but it coincided with developers opening smaller PRs. In other words, AI made devs break work down more so they opened more PRs for the same work. That's still good, because smaller increments are better, but there' no actual increase in coding productivity, and it means the context-switching burden from review work went up e.g developers slowed down in a different area.

The problem that I'm finding in my own work on figuring out the impact of AI is that there's just no reliable way to connect things directly to AI usage. Most of the tooling does things like "The user used AI on the same day that they opened this PR, therefore we'll assume the AI was used to write the code in the PR." In a mature AI-driven org that might be true, but in the rollout phase of an AI experiment it absolutely isn't.

Until there's a good way to fix that gap in the data any measure of AI impact is going to be horribly flawed.

There is no "tiny" nuclear power plant.

On the contrary: https://en.wikipedia.org/wiki/Small_modular_reactor

It's early days but plenty of companies are trying to commercialize designs for them, including Rolls Royce.

Those are people who don't use their computer in the same place. A laptop is great for them. I know plenty of people who only use their computer sat (or standing) at the desk in their home office. And yet they still insist on buying a laptop. It's odd.