danpalmer

Isn't what people pay for with the NYT or other publications, the journalism and editing? Like, the UI isn't why people pay for it.

Given that this seems to just be a re-skin, why a full site that updates hourly, instead of just a browser extension that rewrites the content in-place live?

I'd argue the improvement has come not so much from fixing hallucinations, but more in getting LLMs to successfully run for longer, which lets them more effectively fix their own hallucinations. This is still an improvement, but at the cost of bigger models, with more tokens, running for longer.

I'd be interested to see benchmarks for this sort of thing plotted against cost, power usage, and elapsed time. If these are all sub-linear then that's great, but if some aren't that could be a real problem for LLMs in the long run.

The people I work with outsourcing their research to LLMs are doing wonders. The people I work with outsourcing the line-by-line level of coding to LLMs are hugely productive. It's honestly amazing to see what is possible.

The people I work with outsourcing their thinking to LLMs are struggling, and if I'm not careful they're creating more work for me in review and feedback. The people I work with outsourcing their codebases entirely to LLMs are producing some of the worst software I have ever used.

Assuming the LLMs did actually run the tests, I think this points to the flaw in TDD and LLMs following goals blindly.

Tests are usually not comprehensive. The tests for SQL engines for example aren't intended to provide full code coverage, because they are intended to be re-usable across different codebases. LLMs can therefore easily over-fit to those tests, solving problems in the dumbest ways rather than solving them in the ways humans would.

The human's goal is to produce a working database engine, the LLM's goal is to pass the test suite or implement the spec it has been given.

If this is being announced here, it was probably started at least a few months ago, and at that point funding was likely looking a lot more forthcoming, so it would make sense to try to raise more than is strictly necessary if the price is good. On that basis, it might not be a problematic shortfall.

And as I said, I've never had a case of these where the access was malicious.

The only cases I really get from this are browsers when I try to log in to my router or NAS, firewall software when I ask it about connections, that sort of thing.

"I want to allow the software to talk to devices on my local network", not "whatever it wants".

I've never had a case of this where the access was malicious, and devices on my network have security controls. For me this popup is exceptionally low value.

Is it surprising? Looking at the demo videos I can absolutely see why it didn't get as much attention. Not that it's not impressive, it is, but these are in different leagues.

Apple sold ~250m of those ~1.65bn, so if you work on it being Android only, then foldables are ~1.4% instead of ~1.1%.

For the foldable market to triple in size, Apple would need to sell 40m this year, which if we assume a flat year for overall numbers (likely given rising costs this cycle and the trend of flattening sales anyway), then it would mean the fold taking 16% of all iPhone sales. For a ~2500-3200 USD phone (base model, international pricing), that's not happening.

Why would you think foldables are about to get a lot better? They have improved over the last 10 years, but the original caveats are pretty much all still there, and the iPhone Duo doesn't appear to improve on any of them in any meaningful way.

I'm on Hugo. It's simple, capable, well documented, LLMs understand it well, and I don't have to touch the JS ecosystem. My previous setup was with a Node based SSG and it broke on literally every update and had a hundred dependencies.

I suspect this might be my blindspot. I use my phone a huge amount, but almost never for watching video or gaming. Being able to have a "full screen" video, and another app running "full screen", with the effectively 2x screen space, makes sense.

I had a flip phone and I got it for this reason, but it honestly just sucked. The outer screens are too small to be useful, or in the case of Samsung, just don't have the right feature set to be useful. That means the phone ends up just being a full size smartphone with more steps to be able to use it.

I don't think this is a fair comparison, they are different markets – most people need a smartphone today, nobody really needs a PS5. A better comparison is to say that of smartphones sold in 2025 (~1.65bn), foldables represented ~1%. That's pretty niche.

Nearly 10 years in to foldables, and they still bring so many caveats. They are consistently worse for size, weight, shape, cameras, battery, robustness, and longevity. I don't think I'd ever recommend one.

On the other hand, the advantages are... an iPad like experience sometimes? Or a smaller phone for the flip form factor? The flip form factor certainly does not pay off. I'm honestly not sure who the trade-offs pay off for.

The economy has many different resources with different supplies available. Making the supply of one resource infinite may sound like a huge win, but nothing is dependent on a single resource, everything requires a combination – just like in board game economies, you can't win by only producing wood.

AI represents some weird intelligence-like resource. It clearly does something of value, but that must be combined with other resources. In tech there's a lot of that, so making it (roughly) infinitely accessible has had a nice impact and we've run into our next limiting factor: human time. Companies that aren't "pure tech" though will run into much earlier limiting factors – hardware, compliance, regulation, artistic value, human time, and so on.

Oh no you've completely misunderstood my point. I prefer people, and that's exactly why I want to continue talking to my friends.

Perhaps you're more comfortable chatting to strangers, perhaps this is a cultural thing. An analogy, if I was sitting in a restaurant with friends having a good time and a random additional person got seated at our table that would end the conversation, stop the socialising, and at best we'll have a boring, shallow conversation with the new person about who they are.

Arc was great, I never understood why they moved to Dia instead of just putting AI features into Arc.

The only reason I can figure out is that they decided the technical foundations were just wrong, which would be surprising because they put an astonishing amount of work into building some very specific technical foundations for Arc.

Now that Chrome has native vertical tabs I'm back on Chrome, but I do miss Arc's tab tidying and organisation.

I'm not sure you read my comment, certainly your comment doesn't seem to rebut the point I was making. Cabs are black in my country, but I specifically said "taxi" because this is not about Uber or modern ride-sharing, I've had the same experience with Ubers, black cabs, minicabs, booked taxis, and private drivers.

It's a downside of public transport too. Possibly my reason for being sensitive to this though is that I almost exclusively use public transport, getting a taxi is relatively rare for me, and I don't own a car.

Absolutely! Or even just be alone with your thoughts.

I always felt weird alone in a taxi – having someone serve you that directly in that way, particularly if you're sitting in the back seat, just feels a little uncomfortable. But in a Waymo it's just me.

They make it quite clear in Waymos that nothing/no one is listening until you call them or they have a safety need to interrupt. I've never experienced that, but from those I've spoken to that have, apparently it's very immediately obvious, as it should be, and rare.

The thing I unexpectedly like most about using Waymo is not having a stranger inserted into your conversation.

Every time I've taken a taxi in a group, the conversation ends when we get in and the driver is suddenly a part of our conversation. It's weird to have them sit there listening to your discussion but not participating, and it would be weird for them to participate. Inevitably the conversation turns to the mundane, questions about whether it's a busy day, how often do you drive in this part of town, etc.

I got a Waymo with friends and the conversation never missed a beat as we got into the car, rode for 15 minutes, and got out.

3) Write a piece persuading readers to go and watch it by making the case for how masterful it is at storytelling, without explaining the story.

I'm glad that pieces like this exist, encouraging people to go and see it, and I look forward to reading more of (1) once I've seen it.

> This actually says nothing at all about how the documentary deceives you

That's because the film is easily spoiled, and the impact comes from watching it. Writing a film review without spoiling a film like this is hard.

Or maybe a better way of putting it... a corrupted save annoys one user who has already paid you the $30 they're ever going to pay.

Compared to if you run a cloud storage company, a corrupted file might annoy a user that pays $1m every month, to store data that they themselves make money on, etc.

The stakes for a game save are a lot lower.

I sort of agree, but I can see their point. The author seems concerned that they strongly disagree with Blow, and have a piece of his writing that, if it echoes his world view, might end in a way they don't support, and they haven't got to the end yet.

I'm personally less concerned. Partly because I attribute Blow's social opinions to naivety and ignorance rather than malice – e.g. I think he holds misogynistic views not because he is trying to be misogynistic, but because he's naive about the systemic bias in society, a criticism that can be levelled at many people.

But otherwise, I'm less concerned about Order of the Sinking Star because he has said in interviews that the story is supposed to be interesting to a wide audience and not intended to be based on his world view in any way. The Witness also wasn't some right wing think piece, it was just pretentious.

I get why a journalist would be concerned, but I think they're being overly cautious.

He has said after the game. They wanted to get the game launched, early bugfixes, etc, and then follow it up soon with the language/engine, initially closed source, but eventually open source.

Just a guess based on the way he's spoken about it, but I'd expect an open beta in December-February. I'm not expecting that to be open source though (maybe the engine will be but not the language, unclear).

It's built as a full 3D game, and has extensive use of nice shaders for transparency effects like the crystals, water effects, particle effects, and more. This is all pretty standard game development stuff, you'd be able to do this pretty quickly with an off the shelf engine like Godot, Unity, or Unreal. Godot is a good place to start if you want to learn some of this stuff!

However Order of the Sinking Star is also made in a custom programming language and custom engine. This is not at all normal for game development. There also doesn't appear to be a scripting system. Normally you have an engine built in C++ or something fast, and then a scripting system where you write a higher level language to script how the game runs. Engine development is more separate from scripting development. With Order of the Sinking Star, the game engine appears to be much more just like a library, and the same codebase runs the engine and game logic, with no scripting layer, as far as I can figure out from public hints.

Fair enough, it has definitely evened out now that I've completed the 4 base stories. Reaching the Antipode though has certainly ignited some motivation, I want to figure out the red areas, and getting into the deeper AxB areas is still interesting and I suspect I've got another 20 hours of that to go at least.