ABrockwell @aebrockwell@social.cleverlibre.org
Expert in probability, statistics, and data science, with some programming skills as well.
Former academic and current quant trader.
Joined Nov 2022
@ct_bergstrom They certainly don't have the ability to manipulate humans directly.
But they do have incredible (dangerous) potential as a tool to allow humans to manipulate other humans. People don't seem to have discussed this yet, but anyone (or group) could just build a new #LLM trained with emphasis on different data sources - qanon material, fascist texts, etc. Such new flavors of LLMs could be very compelling and very dangerous to a lot of people.
ABrockwell
boosted
There is a cult of ignorance in the United States, and there has always been. The strain of anti-intellectualism has been a constant thread winding its way through our political and cultural life, nurtured by the false notion that democracy means that 'my ignorance is just as good as your knowledge.
~ Isaac Asimov
@dgar Closely related quote from Carl Sagan:
"The dumbing down of America is most evident in the slow decay of substantive content in the enormously influential media, the 30-second sound bites (now down to 10 seconds or less), lowest common-denominator programming, credulous presentations on pseudoscience and superstition, but especially a kind of celebration of ignorance."
~ from Carl Sagan, "The Demon-Haunted World : Science as a Candle in the Dark" (p. 25):
Some writers and scientists have been ringing the alarm bells for a while now.
ABrockwell
boosted
Elon Musk seems relentlessly, comically hellbent to prove my point on the Myth of the Secret Genius. https://brianklaas.substack.com/p/the-myth-of-the-secret-genius
ABrockwell
boosted
A photographer took a picture just as this couple threw ashes into a river.
He asked who it was.
As they answered their dog.
The photographer asked for permission to edit their photo and this was the result.
ABrockwell
boosted
My colleague and coauthor Jevin West has been warning me for years about the mess that we would face as #GPT and other text-based generative AI matured.
I was not sufficiently concerned.
I just realized why.
I thought of them as tools for deception, and while they offered some advantages for creating disinformation at scale, human beings are quite adapt at creating bullshit on their own.
What I missed was that people would WANT to consult them, knowingly, despite all their issues.
ABrockwell
boosted
Most folks I know who are worried about AI aren’t worried about these programs becoming self aware and oppressing us. That’s not a thing. We’re worried about all the ways these systems will ramp up existing inequalities and further obliterate our shared reality as human beings. That’s already happening.
ABrockwell
boosted
Americans Tommie Smith (center) and and John Carlos (rear), and Australian Peter Norman (front) during the medal ceremony for the 200m dash at the 1968 Olympic games.
National Museum of African American History and Culture, Washington DC.
ABrockwell
boosted
'I have long argued that fact-checking is anachronistic and should be abolished in favor of much more aggressive reality-checking and gaslight-fighting in the main news columns.'
“Fact-checking” is a feeble, inadequate way to respond to racist, antisemitic incitement
@leovarnet Interesting recent discussion of #LLM. A couple of things strike me too on a quick read:
1. There is a big claim in there (based on the Baroni 2022 reference) that essentially "fitted model = theory". To me that feels like a way of not just moving, but actually redefining the goalposts.
2. A lot of the criticism of Chomsky's work is based on weak arguments. Lots of "Chomsky says A but we see B, therefore A is undermined." But as far as I can tell A and B are not necessarily mutually exclusive.
The Chomsky criticism in the paper reminds me a little of the Bayesian/Frequentist conflicts that I use to see when I was in academia (statistics).
ABrockwell
boosted
Recently here has been a lot of excitement around this paper by Piantadosi claiming that Chomsky's approach to language is now refuted by large language models (https://lingbuzz.net/lingbuzz/007180). And I am quite sympathetic to this idea, so I decided to give it a read. But... am I the only only one to find this paper deeply flawed? (1/4)
ABrockwell
boosted
Amazing how many brilliant scientists will look at an LLM "passing" a standardized test and think, "wow, this computer is very smart" and not "standardized tests are very bad at measuring intelligence"
ABrockwell
boosted
@freemo it took me a long time to choose a server! I eventually gravitated to qoto because I liked the description of it, it sounded like a place that respected free speech and was populated by people who largely wanted to do good and discuss their academic work.
Seems like I made a good choice!
@caycepollard Well argued. In my opinion, pretty much the same argument applies to computer programming and more generally, pretty much every field of scientific research.
I think the concept of "Signifiers" can be extended. Another critical thing that's missing is something like "Axioms", that is, unassailable truths. For example, I'm not going to let anyone tell me that 2+2=5 because 2+2=4 is stored in my brain as an axiom, along with many other important things that I know to be true. LLMs embrace alternate facts like these without hesitation.
ABrockwell
boosted
ABrockwell
boosted
Tons of names in my phone are "Kenny Landlord" or "Mike Electric" or "Pete Do Not Answer" and just realized this is how anglo surnames have always worked
ABrockwell
boosted
#ChatGPT and the like aren’t even close to thinking.
“The tendency to view machines as people and become attached to them, combined with machines being developed with humanlike features, points to real risks of psychological entanglement with technology.”
ABrockwell
boosted
No, no, no, God damnit, Washington Post, this is not a "new normal." The point you should be making is just how abnormal this is. But instead you quite literally normalize the criminal running for president.
Trump campaign prepares for ‘new normal’: Running under indictment https://www.washingtonpost.com/politics/2023/03/21/trump-2024-ny-case-republicans/
ABrockwell
boosted
@SarahAnneDipity
This, this, this. EVERY pregnancy is a risk to that person’s life, and no one should be forced to disclose the most personal of personal information in order to try and convince the state that they should be allowed to have full legal autonomy over their bodies. Abortion choice is the pro-life position and we nee to take that term back from the forced-birth activists / religious extremists.
@jr From the article:
When I asked Bard if what it had done was plagiarism, it came clean and even said "I apologize for my mistake and will be more careful in the future to cite my sources."
My thought:
This kind of text (yet again) preys upon the human desire to anthropomorphize. It's an empty promise, because the bot has no intentions, let alone ability, to be "more careful in the future". But a human reader will feel good about this statement if they imagine the bot to be sentient and willing and able to learn from its mistakes.
Extremely slick marketing of a totally empty promise. And the marketing is built directly into the product itself.
Expert in probability, statistics, and data science, with some programming skills as well.
Former academic and current quant trader.
Joined Nov 2022
