eli

Right so that's a trade secrets argument. I agree that's probably their best current legal avenue. It isn't related to ToS though.

There are a bunch of routers like this eg https://openrouter.ai/openrouter/auto

The law was probably fine - that case was just wrongly decided based on the outcome the majority of justices wanted.

What do you suppose the damages would be for a ToS violation? The difference between subscription rates and API rates?

OpenAI accused Deepseek of misappropriating trade secrets which could have serious penalties but seems like an awfully hard case to make.

Seems like we’d all be better off with a law that governs data sharing among AI companies, if that’s the policy goal.

I'm not doubting it's possible to pass such a law, I'm doubting that's it's a practical or worthwhile goal.

The terms of service don't even necessarily matter here. OpenAI could cancel your account for almost any reason, or for no reason at all. They don't particularly need to cite a ToS violation just as a store owner doesn't need to point to a written policy to kick you out of their store.

If the underlying issue is that LLMs should be regulated as a public good, then lets have that discussion. If it's that the major AI companies are becoming too powerful and anti-competitive, let's talk serious anti-trust enforcement. Micro-managing business policies isn't going to work very well.

Seems only fair that if LLMs can use copyrighted data for training then they should be able to use cannot-be-copyrighted output of other LLMs.

But barring the terms of service from forbidding distillation seems like a tough sell. OpenAI shouldn't be allowed to decide what types of customers it wants and doesn't want?

That's the version that goes directly to the trash.

You could just use my.name@gmail.com for the no-alias version. So myname+foo@ works and my.name@ works but myname@ goes directly to the trash.

I think a better way to regulate content moderation would be to pass laws that regulate content moderation. Not backdoor regulations through very broad interpretations of product liability.

HN surely has some underage users. If they look at their analytics and tweak the interface to improve time on site, are they also intentionally creating a product that is addictive to children?

These types of lawsuits seem dangerous. But Meta is a pretty awful company and deserves some sort of comeuppance. I worry about what it leads to though. Hard cases make bad laws.

I'd suggest first looking into the conditions that enable humans to generate sustained, high quality output.

No, you can't just "average" different studies and I'm not sure what "neutral" means in the context of some studies showing a benefit and others not showing a benefit.

The placebo effect is not an excuse to allow drug companies to make false claims about the efficacy of the ingredients

Obviously there are advantages to not having to do work yourself.

But for a benchmark with the goal of picking a model to replace a human on some task? I really think the human should judge which is best.

I haven’t gotten very far yet but I had an idea for a personalized benchmark tool that walks through your git history and helps you craft prompts for tasks that bugs or features already implemented by hand so you can compare how different LLMs would do it.

It isn't.

I actually really like subjective benchmarks, so long as it's a human (ideally me) grading the results. LLM as judge never made much sense.

Fireworks.ai is solid. And if you care more about speed than cost they have a "fast" variant that I think just throws more hardware at the model for about 2x the cost.

Of the accounts involved, yeah. So they can lock them out.

Seems like a pretty straightforward approach to collecting session logs from a bunch of different people/devices would be to have them all set their base url to proxy.deepseek.whatever which logs the data and forwards to the real API.

For being flagged as possibly a competitor? They nuke your account.

Are you comparing the cost of hosted Opus to running Qwen 3.6 locally? That doesn't really seem fair.

I assume it’s just oversubscribed. I’m sure it “can” go faster. But yeah that was my point.

This seems a little fanciful.

There's really no comparison between a model that Anthropic allows Google and Amazon to host with one that has been downloaded hundreds of thousands of times and has dozens of public inference providers.

I'm skeptical of how fast "up to" 750t/s really means. Maybe if they make it extremely expensive so it frees up enough capacity?

GPT‑5.3‑Codex‑Spark currently runs on Cerebras chips and it's giving me around 150t/s. Still relatively very fast, but nowhere near the 1,000t/s they claimed at launch. (Also it's not a very good model.)

That said, I'm super bought in to faster models being better for most use cases than smarter models.

Isn’t that definitionally impossible? If they tell you about it then it’s not a shadow ban.

I don't think Google should also be allowed to remain in charge of Chrome at all but here we are.

Maybe. But also Flash, on the mobile devices of that time that did support it, was a miserable experience. Slow and broken and drained the battery.

How do you know what the founders sincerely believe?

If you were playing a text based game, wouldn't you try a few out?

I imagine there are a fair number of war games in the training data and not so many actual transcripts of internal military force deliberations.

The other "cheating" examples are even worse. It's wild to me that people keep designing benchmarks where the answer is lying around on disk or in the git history. "Hardening" the benchmark with strongly worded prompt instructions is bizarre. There are so many agent sandbox solutions. Why not use one and give it only access to the code it should see?

And I'm not sure how they can rule out other solutions also benefiting from being in the training data, just not reproduced exactly. Seems like it should focus on only CVEs from the last 30 days or something.