cbg0

Very large companies already have in-house council or a law firm on retainer.

> the most important race in human history

This feels like it needs a lot of asterisks. A race implies someone will win it - how exactly is China winning by making those models freely available? "The West" can use that tech, it's not locked away in some vault.

Should the EU not enforce its own regulations? (which apply to all, not just US companies by the way)

That "just not universally available" is kind of the crux of it; all that money in a first world country and many people still fear that calling an ambulance will bankrupt them.

> Taxes and regulations suffocate any progress and no one wants to do anything about it.

We could rephrase this as "the shareholders aren't maximizing their value and the average Joe doesn't have a problem with it".

The anti-EU comments are always the same shape, unable to separate the EU from individual countries and misattributing country specific policies to the entire EU. Only a few countries in the EU have wealth taxes, and it's not the EU who imposed them. Disability pensions are not easy to get and require a lot of medical paperwork and approvals in most EU countries. And what does "so poor economically" even mean?

I wish these benchmarks were more closed off. The next model releases will have a huge jump in scores on this benchmark and it will be pretty much useless.

The world is more than just the Americas.

Probably because it's PEBCAK.

What if the old hires are doing it?

Actually "AI spending" is the new "lines of code" metric.

How are they leaving paying users in the dust? This is their cheapest model which you can already use full-blast even on a $20 plan.

McDonald's might be the largest restaurant chain in Italy but that doesn't really mean much given how many small restaurants the country has with high quality food.

I think you just answered your own question.

Edit: In DeepSWE Sol High scores the same as Fable High for ~1/3 of the cost.

On DeepSWE Deepseek is 54.4% and Luna is 67%

This doesn't seem correct.

Estimated final electricity price for large industrial customers in energy-intensive industries:

USA 50 USD/MWh

China 68 USD/MWh

https://www.iea.org/reports/electricity-2026/prices

I think we've gotten to the point where the question can be asked looking back, not forward: think back to your childhood and how much things have changed since then? If you've lived long enough you can notice these things.

I used to have snow for weeks each winter in my city and now if I get 2-3 days of snow a year it's unusual. Extreme weather events where it rains in one day as much as in one month have become a recurring event to the point we're re-planning sewage systems around these things.

Plenty of people oppose it, about 36% of Germany for one example https://www.dw.com/en/german-poll-majority-for-return-to-nuc...

This doesn't seem novel in any way, given the amount of data available online you can already be targeted by fake bank texts, no need for the free wifi hotspot.

How would your banking credentials leak? Every banking app uses encryption. In fact, I'd wager every app uses encryption nowadays.

Why is something like ublock origin not enough for your needs? Is it more about the principle of not wanting to share any fingerprint across the Internet, or some specific tracking concern?

That 6% could be more than your profit margin, depending on what business you operate. There's nothing saying that fines can't keep coming if your behavior isn't corrected, it's not just one fine then they leave you alone.

The point of fines is to correct bad behavior, not to bankrupt businesses which create jobs.

That's a problem for the future, no need to worry investors with these silly questions.

I would, most of these cybersecurity models have not been made available to the public and the larger models have guardrails in place for certain cybersecurity tasks unless you've been specifically approved.

Do you think it might be overthinking? Try it on medium effort.

It sounds like you just need to correct the plan it lays out to avoid the regression? I'm just looking to debug with you, not defending the model. I've mostly used Opus 5 for code reviews & bugfixes.

I don't understand the need for the snarky comment, LLMs can run the test suite and avoid regressions.

Do you not have unit tests, or how does it introduce regressions? You can tell it how to run the test suite in CLAUDE.md