Grok 3: Another win for the bitter lesson

This article is weak and just general speculation.

Many people doubt the actual performance of Grok 3 and suspect it has been specifically trained on the benchmarks. And Sabine Hossenfelder says this:

> Asked Grok 3 to explain Bell's theorem. It gets it wrong just like all other LLMs I have asked because it just repeats confused stuff that has been written elsewhere rather than looking at the actual theorem.

https://x.com/skdh/status/1892432032644354192

Which shows that "massive scaling", even enormous, gigantic scaling, doesn't improve intelligence one bit; it improves scope, maybe, or flexibility, or coverage, or something, but not "intelligence".

> Sabine Hossenfelder

She really needs to stop commenting on topics outside of theoretical physics.

Even in physics she does not represent the scientific consensus but has some very questionable fringe beliefs like labeling whole sub-fields as "scams to get funding".

She regularly speaks with "scientific authority" about topics she barely knows anything about.

Her video on autism is considered super harmful and misleading by actual autistic people. She also thinks she is an expert on trans-issues and climate change. And I doubt she know enough about artificial intelligence and computer science to comment on LLMs.

> Her video on autism is considered super harmful and misleading by actual autistic people.

I’m autistic and I just watched her video. I found it to be one of the best primers on autism I’ve seen. Not complete, of course, and there’s a lot more nuance to it, but very even handed. She doesn’t make any judgements. She just gives the history and talks about the controversies without choosing sides, except to say that the idea of neurodiversity seems reasonable to her. When compared to most of the discourse about autism, it stands up pretty well. Of course, there’s a lot more I want people to know about autism, but it’s an ok start.

Actually, many autistic people (myself included) would find your statement far more harmful. You assume that all autistic people think alike and believe the same thing. This is false. You try to defend us without understanding us.

Don’t do that.

I suppose there’s a possibility that you’re autistic and found it harmful to you. If so, don’t speak for me.

And she was commenting on an AI’s knowledge of Bell’s Inequality, which is PHYSICS. If she can’t comment on that, who can?

There is a misunderstanding: I did not specify that ALL autistic people think like this. Just that autistic people found it harmful and misleading. There is quite a lot of autistic content creators criticizing the video. It does not mean every autistic person needs to feel the same way.

I am neurodivergent myself but (probably) not autistic. The first time I watched the video I actually didn't think that it was that bad and had a similar reaction to you. But once I started to think about it and educate myself more on the topic I realized how bad it is.

Sure it is not the worst video on autism but it still promotes some really bad ableist views.

Autism speaks it is a horrible hate organization. I don't think there is a spoiler tag here so please skip this paragraph if ableism is triggering to you but there is a video of the autism speaks founder where she talks about how she at one point wanted to kill herself and her autistic child because she couldn't cope with having an autistic child and only didn't do it because of her other non-autistic child. She says that while her autistic child is in the background.

I also didn't know about "aspie supremacy" and why people still use the term "asperger" despite being outdated. Hans Asperger was a Nazi Scientist who is responsible for killing thousands of children. He thought some autistic children might be useful as future scientists for the Nazi regime so assigned them the diagnoses "asperger syndrome" while the other autistic children were to be murdered.

I recommend you watch Ember Green on this topic: https://www.youtube.com/watch?v=vaZZiX0veFY