tmp10423288442

This headline is so vague as to be misleading.

What actually happened is that AT&T and Verizon were penalized for mishandling customer location data by the FCC. They paid their fines, but attempted to contest them because they were levied by an administrative action, without a jury trial (which they claim would violate the Seventh Amendment), drawing an analogy to the case SEC v. Jarkesy, in which the Supreme Court ruled that the SEC could not impose civil penalties without a jury trial.

However, the Court ruled in this case that, since the law in question allowed the companies to appeal their fines to a jury trial, that the penalties are constitutional. This is unlike SEC v. Jarkesy, where the SEC civil fines in question had no provision for appeal to a trial by jury.

Furthermore, the Court clarified that the companies have the option to settle their obligations by either paying the fine, or by not paying the fine and appealing. The companies had claimed that the text of the law required them to pay the fine before their appeal, but the Court disagreed.

It costs a lot more than $18,000 to hire a decent developer, pretty much anywhere in the world. Also using a model is better than another developer in some ways, because there aren't two independent minds trying to work with each other.

No, but you do want Opus-tier models to do desktop and office software automation (think about people who intensely use Excel and the like). Actually those might take even more tokens that coding in a lot of cases. Why do you think Claude Cowork is successful, and why do you think Codex is leaning so hard into Computer use?

Can you name a service that charged companies thousands/seat/month that turned out to be almost or completely useless? There's lots of random services sold to corporates that are not very useful (all the random benefits besides health care, life insurance, and other big-ticket items), but the per-seat charge of those is much smaller.

> "unlimited tokens for every employees we don't even care if it actually ends up being a net positive financially"

That was clearly a short-term trend that would obviously get fixed. Doesn't say much about AI coding as a business model.

There are some objections here saying that some US firms are using Chinese AI providers, but I wonder if any of those are subject to compliance. Large firms that are disproportionately responsible for AI spending are all subject to compliance.

I am too. I've heard of the supposedly correct pronunciation, but I can't bring myself to use it. The "PRY-mer" pronunciation is more common in practice.

I command you to love systemd timers!

(Yeah, they're pretty useful, especially after you get an LLM to write all the boilerplate for you. The boilerplate was the main reason I preferred crontab before.)

Not related - many robber barons went bankrupt in the severe economic crashes of the time, such as the Panics of 1873 and 1893. The Gilded Age continued despite bubbles popping.

> Is the alternative not likely that they would have spent a decade fighting a losing battle over office software with Google and Microsoft? Paper was a great product but the big guys have vertical integration so companies prefer their end-to-end solutions (GSuite etc) and I don't see how Dropbox could have easily overcome that.

Slack, Zoom, and Notion all argue against that. Yes, they have to compete against Google and Microsoft's integrated solutions, but they're good enough that they have held their own. Of course they would be bigger if Google and Microsoft didn't have such products.

It has always existed, but how strictly it’s interpreted (i.e., just cash welfare, or also Medicaid, SNAP, and other means-tested benefits) has shifted between administrations. If you applied during Biden’s administration, I could believe the public charge rule was applied very laxly, particularly because it’s rare to get direct cash welfare in the US these days, and even less for an extended period.

> A human is not reviewing every weld.

On civil engineering projects, I’m pretty sure a human reviews each weld. For mass-produced things, maybe not, although a company would not look good in a lawsuit if they had inadequate inspection procedures which allowed a fault causing injury or death to occur.

> how famous FOSS devs of tools that Google use internally couldn't even get past the resume screen at Google to get an interview

As a former Googler, Homebrew was not ever officially supported at Google, or even particularly recommended, particularly because you were not allowed to store source code on your laptop anyway. Homebrew was definitely not used in any production-critical workflow. It's more accurate to say that some Googlers used Homebrew (I myself used Macports and never encountered any additional friction). Homebrew at that time was also unsuited to anything like Google's scale, so it's no surprise the author didn't get any brownie points for it.

Woz isn't some kind of monk. He enjoys his tech toys that he can only afford because he's rich. He's just bad at managing his money and lost a lot of it through multiple divorces.

Woz should have a lot more money than that for being such a large early shareholder of Apple, so that actually speaks poorly to his reputation as a "successful entrepreneur/investor". Some of the reason why his net worth is below expectations is noble (giving $10m of shares to early employees), but most of it is not - 4 marriages as opposed to Steve Jobs' 1 marriage, an impractical attitude in general, and never having any success after Apple, even as an investor.

What, no comments at all? I guess people are really not surprised.

The eternal truth: money talks, bullshit walks

Mentioned on the story about OpenAI's solving a major Erdos problem: https://news.ycombinator.com/item?id=48212493

HTML is famously forgiving as well - that's the whole reason XHTML failed, because one typo in the latter will make your entire web page fail to render with an error. Markdown is probably a little more forgiving, which mattered more with previous-gen LLMs with small context windows. Any near-frontier model should have no problem generating valid HTML.

Also a lot of LLMs are trained specifically to expect Markdown in their instructions, OpenAI's models in particular (Anthropic expects more pseudo-HTML/XML, but that is different from real HTML/XML).

In the sense that its capabilities were a little overhyped as a quantum leap / unique security threat, and it's so expensive that it's impossible for Anthropic at scale. That hype looked a bit silly when GPT-5.5 matched or even exceeded its capabilities while actually being a model ordinary people could access. But its level of capabilities is only currently matched by OpenAI at this time - the other AI providers, even Google (at least publicly) are quite a bit behind.

It's interesting that ChatGPT Pro is the real deal that can write novel physics or math papers (for certain values of novel), while Claude Pro is crap that, depending on the A/B test, may not even provide Claude Code or at the very least doesn't provide Opus. Shows how LLM naming conventions are currently a mess.

Can you name any instance of OpenAI being as trigger-happy with bans as Anthropic has been in the past few months? Codex may have fewer prosumers, but they've added a lot in that time.

Ironically an analysis that Codex or Claude Code might be able to perform

I’m surprised to hear that she’s only now learning the lesson of group presentations - if you’re even slightly more competent than your group, you’ll be doing all the work and the rest of the group will freeload off of you. It’s not like the freeloaders would have been able to answer follow-ups questions before AI.

There’s always the issue that it might be hard to find competent people that can implement a control network isolated from the internet if it has an internet connection by default.

Why do you trust the UK government won’t do the same?

Still not great if you want a family, because daycare/private school are expensive, and housing in a good school district (to not have to pay for private school) are even more expensive.

They were not photovoltaic solar panels, but solar water heating panels - i.e., running a ton of water pipes on the roof where they're perfectly placed to leak and potentially damage historically-significant artifacts. I'm not surprised they removed them when resurfacing the roof (not right when Reagan entered office, but in 1986).

George W. Bush actually installed solar water heating once again, but on the roof of a maintenance facility, which I didn't know about: https://en.wikipedia.org/wiki/Solar_power_at_the_White_House

5 minutes ago, I was seeing the old version of the page in which Claude Pro included Claude Code. I refreshed and now see that it does not.

No one intelligent should be logging into their personal accounts on their work devices in any case - it's always been the case (at least in the US) that companies can do whatever invasive scanning they want on devices they own.