Octoth0rpe

> Maybe this has to do with the leftward trend of the mainstream press.

What mainstream press outlet has moved leftwards? I can't think of any, and I certainly am interested in knowing which those might be. Inversely, cbs, the ny times, and the washington post have all shifted rather noticeably rightward in the last 10 years.

It really could use a good corporate sponsor and a couple of widely known success stories.

> If they can ship code that matches a spec, why does it matter if they’re using ai or not?

The inability to write fizzbuzz strongly implies their inability to understand what they've shipped. Review is some significant portion of the job. Understanding of the product is also part of the job.

Specs are also in a sense, scaled down, fuzzy, natural language descriptions of a feature. The fuzziness is the source of a bugs, or at least a mismatch between the actual desired feature and what was written down at spec writing time. As such, just matching a spec is just the bare minimum that a good dev should be doing. They should be understanding what the spec is _not_ saying, understanding holes in their implementation, how their implementation enables or hinders the next feature and the next, next feature, etc. I don't think any of that is possible without understanding what was actually implemented.

The difference comes down to focusing on what we think of as the core infrastructure of the internet (the first modern browser, apache), versus _access_. I think there was a pretty big jump around 1996, 1997 in dialup subscribers. To get geekier though, we could actually peg it at late 1993/1994, the eternal september.

I think I'll stick with 1996 though :)

> there are companies that are mostly an investment vehicle and companies that have a strong motivation to make changes in the world.

And there are companies that are mostly an investment vehicle whose leaders spin their self-serving decisions as necessary to make changes in the world. Some of them might even be right! (that their decisions are better in the long run, and that ceding more control to shareholders would lead to better short term outcomes but far worse long term outcomes).

which seem to be out of stock in any case, so probably not a loss for apple.

https://www.apple.com/shop/refurbished/mac

That year range definitely surprised me. In my generation I suspect many of us think of winamp as being just sort of always there, but of course it wasn't. When you add it up, it really only existed for <25% of the history of the web (taking 1996 -> 2026 as the time range, I think 1996 is defensible start for when the web becaome accessible to some portion of people in the first world). Hell, I probably have coworkers who were born after the dissolution of nullsoft (interns anyway).

That's the fun thing about greed, it is rarely satisfied :/

> I see a lot of them on the road so somebody must be buying them.

Two counterpoints: for all the opinionated criticisms, the cybertruck is at least quite noticeable, and thusly you may think that they are a higher proportion of trucks than they really are.

Also, you're far more likely to see them drive around in certain locales due to the cost, so that may introduce additional biases.

> An over-engineered solution (complete with CLI, storage backend, documentation, unit tests) for a trivial problem which that person would've solved by an elegant bash one-liner only 3 years ago.

Importantly, I think AI companies are motivated towards the overengineered solutions as they increase the buyer's token spend. I'm not sure how we can create incentives that optimize for finding the 'right' solution, which may be the cheapest (the bash one-liner). Perhaps a widely recognized but not overly optimized for benchmark for this class of problems?

Pretty damning. Would also be interesting to see the number of commits overlayed. The graph tells a great story about the correlation with MS's takeover, but I wonder if at the same time that uptime went to shit, MS was shifting over large numbers of enterprise contracts to github. That would be a more complete story IMO.

None of which excuses this. Can you imagine someone's reaction in 2017 if you told them that github would be below 90% uptime in 2026? It would be unimaginable.

> That is, a feature can largely be written in one file, rather than bits and pieces all over the codebase.

This seems to be at odds with the goal of token minimization. Lots of small files that are narrowly scoped means less has to be loaded into context when making a change, right?

Throwing out another idea: I wonder if we could see some kind of equivalent of c header files for more modern languages so that an llm just has to read the equivalent of a .h file to start using a library.

On the other hand, I don't think I know any millenials that don't have an extremely overbearing HoA that forbids anything other than a grass lawn.

There are general users of the average SaaS, and there are claude code users. There's no doubt in my mind that our expectations should be somewhat higher for CC users re: memory. I'm personally not completely convinced that cache eviction should be part of their thought process while using CC, but it's not _that_ much of a stretch.

> Torvalds jokingly named it "git" after the slang term, later defining it as "the stupid content tracker".”

I think the better Torvalds quote was when he said "I name all my projects after myself"

Yep

> the US economy really only cares about profit

Which would be ok if we more effectively were able to include externalities into company's overhead, instead of constantly subsidizing them.

Starlink's maritime, roving, airplane, and military options are all much more than $100/mo/user. Not sure how much that closes the gap, but it's _something_.

Source: https://starlink.com/business/aviation ($250->$10k/mo)

https://starlink.com/business/maritime ($250/mo)

https://starlink.com/business/mobility ($65->$540/mo)

Does it make sense to value Starship Commercial Launch at $170B, _and_ Falcon 9/Heavy at $100B? I would expect that if Starship achieves its operational goals, then it should quickly deprecate nearly all uses of Falcon, the exceptions being national security launches that require validating the launcher, or Dragon launches for similar reasons. Even those categories are likely on a countdown the moment starship is rapidly reusable.

> it can take anywhere between 6-25 seconds for a response (after lots of thinking) from me asking "Hello world".

That's not an unsurprising result given the pretty ambiguous query, hence all the thinking. Asking "write a simple hello world program in python3" results in a much faster response for me (m4 base w/ 24gb, using qwen3.6:9b).

I tend to agree.

> I see no reason for the US to pay for bases it's never going to be allowed to use.

Which isn't the situation being imposed by Spain. They're being told they can't use the airspace for one specific military action. They maintain use of their bases in other ways (training, presumably ship refueling, maintenance, etc). They may be able to use the airspace for _other_ military actions in the future.

I hate to say this, but they're correct, if only pedantically. The claim was:

> The United States has never activated article 5

The US didn't activate it. It was:

> The decision to invoke NATO's collective self-defense provisions was undertaken at NATO's own initiative, without a request by the United States

source: https://en.wikipedia.org/wiki/2001_NATO_Article_5_contingenc...

Regardless, article 5 was activated _on behalf_ of the US, if not at the US's request.

> It's kinda crazy we allow countries like Spain

It's crazy that we recognize country's right to determine the use of their own airspace??? Are you serious?

Cool. Now do all government offices / properties of any kind please (and also go national with the policy).

Absolutely fuck these things and anyone who advocates for them. No exceptions.

> reasonably affordable and available smart glasses have finally begun catching on within the last year.

Also, no they haven't.

With only a little sense of self aware irony, one thing I hate about so much dialog these days is how vehement opinions are. I don't particularly like the rounded corners, and think it's a regression. It's also... fine. It's not the difference between usable and entirely unusable. And I see this kind of attitude all over the place now. A slight change, some slightly non-ideal behavior and all of a sudden a product is THE WORST THING EVER. We will be ok with inconsistently rounded windows. I think people need to be a bit more tolerant of design decisions that are opinionated, and likely worse but also not breaking.

Ads in a start menu can die in a fire though.

> A single patched llama-server runs on K3s, providing both generation with speculative decoding (~100 tok/s)

There seems to be at least some detail on that point.

And notably, that page makes this post's title inaccurate. As of this morning, it says `90.21% uptime`, which is a _single_ 9, not 3 (though that's for the platform as a whole, no individual component appears to achieve three 9s.)

Even if that happened tomorrow, I suspect we'd have _at least_ a decade of people tweaking/optimizing designs on the same node to squeeze meaningful performance upgrades out. Eg, coming up with hardware support for new int/float formats that make more sense for the models of 2029, running matrix operators on ram chips directly, etc.

> pretty soon you can buy one big ass server that will last potentially decades as it would be purpose built for ai.

This feels like a very, very weak prediction (though certainly possible).