I like AI, but seriously, who actually invests on this basis? Where is the critical thinking? I don't feel sympathy for any investor that gets rug pulled on this stuff.

Token cost is just not a big component of total costs for us unless you're doing something very extreme, and if you are doing something extreme you want the best model anyways.

No, the opposite. In my experience immigrants care far more about getting good grades, whereas most multigenerational American students were happy with Bs or even Cs.

> What is the specific ethnic breakdown of the set of people you consider to be top AI researchers at the top labs? How does this compare to 1) the current ethnic breakdown of the totality of the United States of America, and 2) what the ethnic breakdown of the United States of America would be under your preferred immigration policy.

A lot more Asians. Very few Asians. A lot more Asians.

> What kinds of immigrants have you met, and not met? How many of them can you talk with in the language they are fluent in, in order to get an accurate sense of the degree to which they embody the American spirit?

Chinese, Japanese, Korean, Russian, Nigerian, Mexican, etc. So many.

The only ones not fluent in English were the Hispanic immigrants, but despite this they better embody the American spirit than most Americans. I don't need to be fluent in Spanish to see that (though mine is passable).

The skilled first and second generation American immigrants do extraordinarily well. Most of my second generation Asian peers are clearing mid 6 to low 7 figures in their 30s, many working on their own ventures or at bold startups. And my Hispanic landscaper that came here with nothing, now owns a business enough to pay him and his four employees.

Now compare this to the median multigenerational American - working a dead-end job, comparatively far less grit, ambition, and risk-taking, too comfortable so there is not as much a drive to be exceptional or prove themselves.

Which group do you think the Founding Fathers would say better reflects the American spirit? To me immigrants are clearly the better reflection of the best aspects of American culture.

Go to any high school and see how little American schoolchildren care about academics vs immigrant schoolchildren. Academic excellence being uncool is baked into American culture. You're a "nerd" if you do well and care. Getting a B is "good enough." And "C's get degrees." This mentality is plainly unacceptable in most immigrant cultures.

I took almost two dozen AP classes in my day. In each one, the concentration of immigrant groups was far higher than the rest of the school at large.

Expand this out to college. Look at the admissions for top colleges without affirmative action. How do their demographics compare to the rest of the country? - MIT, 47% Asian. Berkeley, 41% Asian. UCLA and Stanford, 27% Asian.

6% of the US population is Asian, and 75% is white, and these schools don't have affirmative action. If all groups were equally competitive, admissions would reflect demographics.

(Ethnicity here is a crude approximation for immigration recency. I am not saying one ethnic group is better than another - simply that children of immigrants excel.)

The same goes for top PhD programs, the highest paying STEM jobs, even C-suite positions at big tech.

I am American and when I say that we have a problem where most Americans do not give a fuck about education, I am not being racist, I am just pointing out the truth. Over decades, our culture has bred an anti-intellectual attitude, one that prioritizes being cool and sociable over getting shit done. This is the antithesis of progress and ambition. It is great for sitting around and demanding handouts.

Immigrants more closely approximate the culture the founding fathers intended for the US. They uproot their lives to build something great. They get off their asses, do exceptionally well, and are carrying the nation on their backs. The rest of our culture could learn from them, instead of blaming our problems on them and turning them away.

This isn't hard to understand

"Model is nerfed" claim hits social media

Someone else sees it, frequency bias makes them think their model is also nerfed, and they amplify the claim

Now it spreads, like a virus, even if the model never changed

Social dynamics like this are well understood psychologically

> If the LLM used to be able to achieve set goals and no longer could, it is already a sign of the distribution shift.

The more likely explanation is that you're looking at older LLMs with rose tinted glasses, and misremembering what it could achieve

Otherwise you could measure the token shift and see the better tps and latency

Your own evals would trend down

But no one, not one person, has presented empirical evidence of being served a quant. Just vibes.

> This is data.

Nope. Because frequency bias is a thing. If you hear on Twitter "model X got nerfed," your brain will look for that pattern and notice it more than usual. This will then confirm your suspicion, which leads to a vicious cycle. Then you tell your friends and the same phenomenon repeats.

None of this requires the model to get worse. It's a well understood psychological phenomenon.

> I can tell you what it means: models performing worse at coding tasks. So people report models being worse at coding tasks

The perception of a model performing worse at some coding task is not what "different token distribution" means. You should ask AI to explain my comment ;)

Latency and TPS can also tell you if you're getting a quant.

Anyways you should really get some help. Praying for you!

I think there is a smaller argument that the newer Kindles don't feel as nice. The Oasis was the pinnacle of e-reader hardware design, and it'll be sad when they stop supporting it, but it certainly won't be worthy of a news article or this kind of reaction.

When agents write most of our code, I question if we will still even need documentation.

People around you can "smell" your passion and sometimes it energizes your team. It makes people around you give more of a damn.

You really just have to find something you care about. This is especially easy at the big tech companies, but for some reason, most engineers don't even think about it - they get stuck in this miserable loop of stress and hating their work.

For agent written code I now default to no documentation and explanatory function signatures, it works better for me at least.

> There is data

Please show it.

> while ignoring the actual meaning of the words

It was an incoherent mess of insults, so I am still not sure what you're trying to say.

> Yeah, in your dreams maybe

So now I'm lying about my employment on an anonymous forum for... what, exactly? If you are actually this conspiratorial IRL, get help.

No, feelings are not reliable data when frequency bias and misinformation exist. There is a reason most experiments isolate out bias as much as possible.

> Moreover, benchmarks (that matter) are not run continuously all the time.

So there's no data?

> What do you think different token distribution means?

You clearly did not understand anything I said. Stated simply: If you were being served a quant, you'd be able to tell by looking at the token distribution, latency, and TPS. You don't need to trust the labs' word for it.

> they pay you your salary, duh.

In fact, I get paid by a FAANG, though I do use Anthropic products heavily. Further, I don't really need money, I have more than enough. So much for reading my history.

> You are actively poisoning this forum

Your degenerate discussion - calling people shills instead of engaging with the argument, insulting them when your arguments are disproven, your inability to hold a rational debate that's not angry and emotionally charged - that is what is poisoning this forum.

Frankly, if you react this angrily and emotionally to a simple rational premise (that frequency bias leads to the perception of models being worse than them actually being worse), you're ngmi unless you're already independently wealthy.

I would recommend a therapist, it helped me when I had similar behavioral issues. (Claude is a great therapist, by the way ;)

Most of them. We have normalized getting Bs and Cs in our schools. Our school curricula are mediocre, and our culture around education is as well. It is distinctly uncool to care about education here.

> Which specific immigrants are smarter than the average American and are therefore responsible for the nice things about America?

Most of our best doctors, scientists, and engineers are all immigrants. Look at the ethnic breakdown of top AI researchers at the top labs.

> which isn't specifically gated on the intelligence of individual immigrants will improve a country along this metric.

It's not just intelligence. Immigrants overall have more grit, more entrepreneurial spirit, and more ambition and willingness to succeed than median Americans. It takes a lot to uproot your life and attempt to make it elsewhere. The vast majority of immigrants I've met embody the American spirit far better than most born-and-raised Americans I've met.

> And in fact the US has a huge number of legal pathways for immigration

That we are making harder and needlessly painful, which will in turn reduce the amount of highly intelligent and capable immigrants we get as well.

TBH most immigrants I've met better embody the American spirit than most Americans.

> so the anecdata does gradually become data.

No, it does not. Countless social phenomena demonstrate how factually incorrect misconceptions spread rapidly. Frequency illusion is real and contagious.

> How do you know that [they are not serving quants]

Lots of ways to tell, if you weren't busy calling people shills.

First, Anthropic and OpenAI have both stated they don't serve quants. Weak protection, but it's there.

Second, no one has shown an A/B or eval proving a regression.

Third, and most importantly, the actual output measurably changes. Quants have a lower latency, higher TPS, and different token distribution. Despite having access to this data, no one has any evidence proving a quant has been served.

> You are an Anthropic shill

I'd explain the reasons I favor Anthropic over the others, but you'd just go back to yelling "shill" instead of engaging in a real conversation. That said, I am a fan of GDM as well, and think Gemini is better than Anthropic for everything other than code.

I've seen nothing resembling sane, reasoned thought from you in this thread. Just anger.

You haven't substantively debated a single point, it's like "shill" is the only word in your vocabulary. Again, this isn't Reddit.

I do not understand why the "American First" MAGA crowd can't get it through their thick skulls that everything nice they have, including our technological lead, is built by immigrants that are just smarter than they are.

This is just an ego problem I suspect. It bruises the ego of MAGA voters to realize that immigrants actually are smarter, they actually do get paid more (and not because they're "taking the jobs" but because they are actually more desirable.)