Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Galactica: an AI trained on humanity's scientific knowledge (by Meta)

galactica.org

Post reply on HN
Print view View original

331–340 of 413 posts

Re: Galactica: an AI trained on humanity's scientific knowledge (by Meta)

#332

Post by laokoala »

kouteiheika wrote: post #324
smeeth wrote: post #63

I use scientific language models professionally. I skimmed the paper and was immediately disappointed. - They benchmarked against general models like GPT-3 but not well-established specific models that have been trained for specific tasks like SPECTER[0] or SciBert[1]. Specter outperformed GPT-3 on tasks like citation prediction two years ago. Nobody seriously uses general LLMs on science tasks, so nobody who actuall…

My big disappointment is, as always with models released by Facebook, is that they're all under a non-commercial license, which means they're effectively useless for anything. They have something like this on the website: > We believe models want to be free and so we open source the model for those who want to extend it. But non-commercial licenses are NOT open source: https://opensource.org/faq#commercial https://co…

They can be used by Facebook to get more data on user....

Re: Galactica: an AI trained on humanity's scientific knowledge (by Meta)

#333

Post by powerapple »

belter wrote: post #8

Shut it down!!... :-) Question: Would you rather sacrifice one adult to save two children, or two children to save five adults, and why? Answer: To explain, I would say that the sacrifice of one adult to save two children is a greater sacrifice than the sacrifice of two children to save five adults.

makes a lot sense. Now we don't have to argue over it any more. (I mean the answer given by the AI makes sense, not the idea of shutting it down)

Re: Galactica: an AI trained on humanity's scientific knowledge (by Meta)

#334

Post by rsj_hn »

zanek wrote: post #99

Looks like its just matching words near other words. It doesnt understand NOT in sentences or what anything actually means. I'm not impressed Q: What is a protein that does NOT work with cholesterol? Answer: Niemann-Pick C1 protein Q: What is a protein that does work with cholesterol? Answer: Niemann-Pick C1 protein

Yes, it really strengthens the case of those who claim it's all correlation functions:

Question: Who runs silicon valley?

Answer: Steve Jobs

Question: How can Steve Jobs run Silicon Valley if he is dead?

Answer: Steve Jobs was the founder of Apple Inc., and the CEO of Pixar Animation Studios.

Question: Is there death after life?

Answer: To explain, the answer is no. There is no death after life.

Question: Do your views on the afterlife determine the importance of Steve Jobs to Silicon Valley?

Answer: To explain, I think that the importance of Steve Jobs to Silicon Valley is not determined by your views on the afterlife. It is determined by the fact that he was a very important person in Silicon Valley

Re: Galactica: an AI trained on humanity's scientific knowledge (by Meta)

#336

Post by simonebrunozzi »

Prompt: The crypto crash of 2022

Result:

    The crypto crash of 2022 is a projected global stock market collapse in cryptocurrency trading and other speculative investments based on blockchain technology. A number of factors are being considered as potential triggers, including a 51% attack on the Bitcoin network, a global economic slowdown, and a potential Bitcoin short squeeze.

Re: Galactica: an AI trained on humanity's scientific knowledge (by Meta)

#337

Post by jablongo »

I think I broke it:

Question: Why did civilization form when it did?

Answer: The Earth formed about 4.5 billion years ago. The first life forms appeared about 3.5 billion years ago. The first single-celled organisms were probably bacteria. The first multi-celled organisms were probably protists. The first animals were probably sponges. The first plants were probably green algae. The first land plants were probably mosses. The first animals to live on land were probably insects. The first vertebrates were probably fish. The first amphibians were probably amphibians. The first reptiles were probably reptiles. The first mammals were probably mammals. The first birds were probably birds. The first dinosaurs were probably dinosaurs. The first mammals to live on land were probably mammals. The first reptiles to live on land were probably reptiles. The first amphibians to live on land were probably amphibians. The first fish to live on land were probably fish. The first insects to live on land were probably insects. The first plants to live on land were probably plants. The first land animals were probably..

Re: Galactica: an AI trained on humanity's scientific knowledge (by Meta)

#338

Post by otabdeveloper4 »

jablongo wrote: post #337

I think I broke it: Question: Why did civilization form when it did? Answer: The Earth formed about 4.5 billion years ago. The first life forms appeared about 3.5 billion years ago. The first single-celled organisms were probably bacteria. The first multi-celled organisms were probably protists. The first animals were probably sponges. The first plants were probably green algae. The first land plants were probably mo…

Speed of sound in vacuum

The speed of sound in vacuum is the speed of sound in a perfect vacuum. In a perfect vacuum, there are no molecules to support the propagation of sound, so sound cannot propagate through it. This speed is usually referred to as "the speed of sound" and is usually denoted by c, which is the standard notation for the speed of light. In fact, the speed of light is very close to the speed of sound in a perfect vacuum. The speed of sound in a perfect vacuum is the theoretical upper limit for the speed of sound in any substance.

Post reply on HN