When you are ready to record a change, split your current working changeset and pick the things you want to commit. It’s equivalent to staging then committing with Git.
As far as I can see out-of-order streaming is only half the described functionality – there is also HTML streaming & revamped DOM parsing which does not have the positive signals that out-of-order streaming does:
Yea we've been working with Mozilla and Apple on that one as well but they haven't responded to the standards position yet. I am sure that we'll reach something that's within consensus.
It’s a brand new market that they want to claim a share of. I doubt they would be making much money of selling deepseek inference right now even if it were profitable, so why not throw sum subsidies at it for a little while in the hope that you are one of the big names left standing once everyone runs out of money.
I don’t know enough to be certain either way. But I will say that I know that Amazon has operated certain product segments at a loss before. Whether that’s with direct price subsidies or credits is irrelevant in the face of a new product with hype unlike anything I’ve ever seen in over 20 years in the industry. It’s highly plausible in the face of this absolute mania and FOMO that Amazon is operating open source inference at a loss to gain market share. They might think that inference prices will drop in the future.
They might be panicking because they don’t have good models of their own. Or they might just be price matching other open source inference providers. They have cut prices to keep up with competition many times over the years.
Whether they are doing it or not, you don’t know they aren’t, and it’s plausible that they are. So the claim that starts with “we know that people are making a profit selling open source inference at X price therefore Y” is unfounded.
There’s a pretty clear example of Zitron being dishonest here:
> I also severely doubt that Anthropic managed to make the cost of running its services profitable in the space of six months.
> [Per The Information in January], Anthropic missed on its gross margin projections, saying that its inference costs were 23% higher than the company had anticipated.
> How did Anthropic, which faced a massive influx of new business to the point that Anthropic was forced to buy more compute from Elon Musk, magically become profitable? Other than that discount, of course.
If you follow the link to The Information, you’ll see that it’s a paid article with the headline “Anthropic Lowers Gross Margin Projection as Revenue Skyrockets”. But what happens when you actually read the article?
> Anthropic last month projected it would generate a 40% gross profit margin from selling AI to businesses and application developers in 2025, according to two people with knowledge of its financials. That margin was 10 percentage points lower than its earlier optimistic expectations, though it’s still a big improvement from the year before.
So, according to Zitron’s own source, Anthropic are actually earning 40% gross profit margin on inference, and that is a dramatic jump upwards! This totally contradicts his position that it’s an implausible “swindle” for Anthropic to claim profitability. He’s counting on the fact that most of his readers don’t subscribe to The Information and will only see the headline, or that they will just see a citation and trust that it backs him up without checking.
According to Dario Amodei, Anthropic are even profitable when including inference as long as you look at it on a per-model basis; it’s just that every model is more expensive to train than the last one.
For instance, if you have already spent $n to train a model and are currently earning $2n selling inference with it; but are concurrently spending $3n training the next model in anticipation of earning $6n with it, then you are already in the hole for $n and are currently also losing $n – but you are doubling your money with each model because your $n investment in the first model returns $2n and your $3n investment in the second model returns $6n.
> can you find the members of this culture that like getting slopgrenaded? A communication culture needs both speakers and listeners. I see the speakers, I have yet to see the listeners. I could just be missing them though.
Head over to Reddit and look for the obvious AI-generated engagement bait. There are usually a bunch of people earnestly responding without even realising they aren’t talking to a real person. Sure, some of those are also bots, but a huge number of them are real people too.
Average people aren’t good at noticing this stuff. They might not deliberately seek it out, but when they are exposed to it they like it enough to voluntarily spend their time on it.
> In the case that the listeners are greatly outnumbered by the speakers
I think it’s quite likely that the people who appreciate receiving slop outnumber the people who don’t.
> I'd like to understand why I can't use a song in one of my videos without permission/payment, but an AI company can train models using that song without having either.
Because when you say you are “using” the song, what you mean is that you are distributing copies of the song, which is protected by copyright.
When AI companies train on the song, the model is learning from it. Outside of the rare cases of memorisation, this is not distributing copies and so copyright doesn’t have any say in the matter.
Learning isn’t copying, so copyright doesn’t get involved at all.
I appreciate your comment, but you answered as if this question had been answered legally. It has not.
The New York Times is suing both OpenAI and Microsoft for copyright infringement. The Authors Guild is suing OpenAI. Getty Images is suing Stability AI. Disney is suing Midjourney. Universal Music Group and Sony have filed suits against multiple AI companies.
> so copyright doesn’t get involved at all.
The dozens of ongoing cases that discredit that statement.
Which statement of mine do you think is not settled law? Which law do you think is being broken and how?
Your objection doesn’t make sense. In the event that an AI company loses a lawsuit for copyright infringement based on simply training on copyrighted works, the answer to you saying you’d like to understand why they can do it and you can’t is simply “your premise is wrong; neither of you can”.
> Which statement of mine do you think is not settled law?
I object to your statement that "copyright doesn’t get involved at all" when that is objectively untrue. If that was true, many of the world's largest companies wouldn't be spending tens of millions of dollars to have that question answered in court. Go to any law-focused forum, and you will find attorneys arguing over these questions.
To train a model using a book, you must first obtain a copy of that book. Did OpenAI purchase a copy of every book not already in the public domain used during training? They did not.
Some of the suits I mentioned claim that OpenAI literally stole copies of books to train its models.
My point is that the copyright question has not been answered. If the NYT, et. al. win, it will be a watershed moment for how AI companies pay for training data moving forward.
It’s not necessarily going to be Google, but the rise of AI does not look good for the web, and it’s a largely self-inflicted wound.
Have you not noticed that the typical user experience on the web is dire? You need to click through tracking consent forms, subscription overlays, put up with dark patterns, etc. Remember, half of all users don’t even use an ad blocker. We’ve collectively made the web a very unpleasant experience.
Along comes a new technology that lets you just say what you want and it will go and find the answer or do what needs doing for you without any of that crap. Of course users are going to prefer it to the crap we dump on them via the web! Can you blame them‽
> Along comes a new technology that lets you just say what you want and it will go and find the answer or do what needs doing for you without any of that crap. Of course users are going to prefer it to the crap we dump on them via the web! Can you blame them‽
The web used to be like that, but then it was enshittified. The same thing will happen to consumer AI, and it will be done by the same people.
You can quibble about the exact numbers, but I think it’s fairly clear at this point that inference is profitable with decent margins. Like you say, unit economics are more interesting than the profitability of the company as a whole.
reply