←back to thread

Grok 3: Another win for the bitter lesson

(www.thealgorithmicbridge.com)
129 points kiyanwang | 10 comments | | HN request time: 0s | source | bottom
Show context
bambax ◴[] No.43112611[source]
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".

replies(7): >>43112886 #>>43112908 #>>43113270 #>>43113312 #>>43113843 #>>43114290 #>>43115189 #
cardanome ◴[] No.43113843[source]
> 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.

replies(8): >>43113905 #>>43114246 #>>43114537 #>>43115760 #>>43116614 #>>43116644 #>>43117944 #>>43121775 #
Mekoloto ◴[] No.43113905[source]
Your statement is missleading.

She doesn't say she is an expert on trans-issues at all! She analyzed the studies and looked at data and stated that there is no real transpendemic but highlighed an statistical increased numbers in young woman without stating a clear opinion on this finding.

The climate change videos do the same thing. She evaluates these studies discusses them to clarify that for her, certain numbers are unspecific and she also is not coming to a clear conclusion in sense of climate change yes, no, bad, good.

She is for sure not an expert in all fields, but her way of discussing these topics are based on studies, numbers and is a good viewpoint.

The funding scam you mention is a reference of "these people get billions for particle research but the outcome for us as society is way to small"

replies(3): >>43114307 #>>43115852 #>>43120600 #
cardanome ◴[] No.43114307[source]
Having studied physics does not allow you to evaluate studies in completely unrelated field in any meaningful way.

Especially not in such politically-charged fields that require deeper knowledge about the historical context, the different interest groups and their biases and so on.

Her video on trans-issues labels people that advocate for the rights of trans-people as "extremists" and presents transphobic talking points as valid part of the scientific discussion.

Her trying to appear "neutral" and "just presenting the science" is exactly the issue. Using her authority as a scientist when talking about topics she has no expertise in.

Here is a debunking of her video on trans-issues: https://www.youtube.com/watch?v=r6Kau7bO3Fw

Here is a longer criticism of her video on autism: https://www.youtube.com/watch?v=vaZZiX0veFY

replies(4): >>43115225 #>>43115696 #>>43117192 #>>43124481 #
bflesch ◴[] No.43115225[source]
So where does your "scientific authority" come from, which is needed before criticizing someone according to your own logic?

You're not even using your real name here. Nobody knows if you have any scientific qualifications, or a university degree at all.

replies(1): >>43115658 #
1. cardanome ◴[] No.43115658[source]
I am writing here as a random hacker-news user. I don't claim to have any authority.

Sabine Hossenfelder presents her opinions as "what the science says" and that is the problem.

replies(3): >>43116154 #>>43116737 #>>43117770 #
2. pyinstallwoes ◴[] No.43116154[source]
She can’t state her opinion? lol.
replies(2): >>43116449 #>>43116530 #
3. fx1994 ◴[] No.43116449[source]
random Internet user, yes, Sabine... no, it's all hyped, you know sometimes you just have to come to your own conclusions
4. ◴[] No.43116530[source]
5. dimal ◴[] No.43116737[source]
She does not. She specifically speaks against that kind of thinking. Recently, she stated it as, “Believe arguments, not people.” I couldn’t have said it better.

She makes arguments, forcefully. That’s good. That’s what science is supposed to be. I don’t agree with her on everything, but I find her arguments engaging, and sometimes convincing, sometimes not. But her process is not dogmatic, as you’re trying to make it out to be.

6. anonym29 ◴[] No.43117770[source]
HN mod dang, if you are reading this, I have a question. I was previously given a warning for a post that levied factual criticisms about the quality of source code contributions performed by a woman who had intentionally put work forward into a highly public open source project in her own name.

I had specifically mentioned her by name in my criticisms, and I was given a written warning that doing so went against HN's policy on "targeted attacks" or "targeted harassment" or something similar.

Why is it okay for this user to suggest that the act of this woman presenting her work publicly "is the problem", while it is a HN AUP offense for me to criticize the quality of the source code contributions written by another woman presenting her work publicly?

I'm not requesting enforcement against this user or a retroactive removal of my warning, I'm just trying to understand the difference better to improve the conformance of my own discourse to the intent of HN's AUP.

replies(3): >>43118486 #>>43118555 #>>43122948 #
7. srid ◴[] No.43118486[source]
In my understanding, one reason Sabine gets readily attacked (as user `cardanome` does here) is because of her criticism of orthodox physics theories. She has famously exclaimed that all cosmological theories supposing a “beginning” of the universe are essentially “a creation myth written in the language of mathematics”.

https://news.ycombinator.com/item?id=32618719

replies(1): >>43120687 #
8. latexr ◴[] No.43118555[source]
It is extremely unlikely Dan is reading that, or the posts above yours. HN mods are only human and can’t see everything, that is why members have tools like flagging.

If you want to contact Dan, email HN and make your case. The concat information is at the bottom of the page.

9. jiggawatts ◴[] No.43120687{3}[source]
Read the linked blog post by Sabine again. That’s not what she says at all.

She’s saying something much more specific about the earliest moments (milliseconds!!) of the history of the universe, and before that time, of which we have practically zero observational data, and can’t ever visit even in principle.

She’s arguing against woo, against science fiction, against unsupported what-if musings that are fun to talk about — but are not science.

10. starspangled ◴[] No.43122948[source]
Interesting question. I'm not sure what "woman" has to do with it since they're both women, but we'll go with it. It would be helpful if you could link your comment, but I guess it's been nuked. Anyway...

Just because you can tie a person to work they have performed using public records does not seem like it should put them on the same level as someone who communicates with and creates work directly to the public, or a public figure. Not even if some of the actual work itself is performed in some open and observable space -- For example I don't think one has any more or less moral right to commentate on and publicly critique the work of a carpenter working on a building scaffold that's easily observable from the public street, than one does about a programmer working on their own idea from their own home in private. That seems like the immediate obvious difference between the two situations you describe. They don't sound equivalent at all, so I don't think you can win your case on that angle.

But work by "non-public-figures" is frequently posted about and commented on at Hackernews. Obviously open source work is a significant source of such discussion simply because it is accessible. Therefore, clearly it's not entirely verboten to talk about that. Is it permitted to criticize? I don't have a particular example at hand but I'm quite certain that I've seen negative opinions about people's work on this site from time to time. I think this is the angle you could argue your case. Was it fair and consistent that yours was called an attack or harassment? Are similar criticisms of work by non-public-figures permitted on here? Without the full context we can't answer that.