Jedesmal wenn ich heute #Bonanza gucke, und „Pa“ aka „Ben Cartwright“ aka Lorne Greene auftritt, denk ich so: ahhhhh, das hat also Commander #Adama gemacht, nachdem die #Galactica die Erde erreicht hat.

#dwr #foto #fotografieren #mywork #goodmorning #fbg #fbd #jamendo #CC #fedibikes #MdRddG #MdRzA #Frühstück #Kakao #Tee

#Tousled #Crane on #Tour

#Welt! Bist du noch da?

#Elbtunnel

Erinnert mich ein wenig an die #Startrampe der #Jäger bei #Kampfstern #Galactica. Nun, da wir #Montag haben, starten wir mal mit 'nem #Pott #Kaffee in die neue #Woche.

https://www.jamendo.com/track/1281723/starship

Bleibt senkrecht und gesund!

#meta #Galactica https://www.thedailybeast.com/metas-galactica-bot-is-the-most-dangerous-thing-it-has-made-yet #ai

🤖🥳 Everyone is obsessed with #generative #AI these days, and it feels like every week there’s a new state-of-the-art model like #ChatGPT, #Galactica, #StableDiffusion, or #Midjourney. My team’s work in #human-centered-ai shows the tremendous value of these kinds of artifact-grounded interactions.
1/8

Blog post about avoiding hallucinations in language models using prompt chaining: https://alanbuxton.wordpress.com/2022/11/30/large-language-models-hype-and-prompt-chaining/ #galactica #flant5 #gpt #chatgpt #ml #nlp

Browsing the papers presented at #neurips2022, it is not humanely possible to read all of them, or even to go through all the abstracts. A tool like #Meta #AI #Galactica would be very useful (if it works) to synthesize the key findings, methods, results for us.

I worry that a lot of the discussion on #Galactica is too focused on classic academic fraud (plagiarism, falsifying data, etc.) We already have ways of dealing with that. I’m *much* more worried about how a more creative bad actor could use LLMs for a DDoS amplification task on volunteer peer review. How many fake papers would need to get past desk reject to bring CHI to its knees? Even just in units of characters typed, prompts are way shorter than the 4 reviews needed to reject the work.

New Meta AI demo writes racist and inaccurate scientific literature, gets pulled https://arstechnica.com/?p=1899005 #largelanguagemodel #machinelearning #Galactica #AIethics #Biz&IT #MetaAI #AI

3000 BC at the #Meta #AI Headquarters #Galactica #largelanguagemodel #machinelearning #humor

RT @Michael_J_Black@twitter.com

I asked #Galactica about some things I know about and I'm troubled. In all cases, it was wrong or biased but sounded right and authoritative. I think it's dangerous. Here are a few of my experiments and my analysis of my concerns. (1/9)

🐦🔗: https://twitter.com/Michael_J_Black/status/1593133722316189696

RT @Michael_J_Black@twitter.com

I asked #Galactica about some things I know about and I'm troubled. In all cases, it was wrong or biased but sounded right and authoritative. I think it's dangerous. Here are a few of my experiments and my analysis of my concerns. (1/9)

🐦🔗: https://twitter.com/Michael_J_Black/status/1593133722316189696

I just asked Meta's Galactica.org to write about the "political economy of Facebook." It returned, "Sorry, your query didn't pass our content filters."
 
Upon checking further, I can't get #Galactica to write anything about Facebook or Mark Zuckerberg.

Not that I was expecting a blistering critical analysis, but it feels pretty weak that Galactica will write about other platforms and tech CEOs, but seems to filter any attempt to write about Facebook or its CEO.

Interesting AI tool: Galactica.org

It can summarize academic literature, solve math problems, generate Wiki articles, write scientific code, annotate molecules and proteins, and more...

#AI #Galactica #DeepLearning #PapersWithCode

https://twitter.com/paperswithcode/status/1592546933679476736

"Sorry, your query didn't pass our content filters. Try again and keep in mind this is a scientific language model." #Galactica

Looks like you can get AI to write literature reviews now. Currently reads like an essay submitted by a first year undergraduate student that doesn't know how to reference.

#Galactica

Q&A for science: #Meta #AI's Large Language Model #Galactica
a large language model #LLM that can store, combine and reason about scientific knowledge. "We train on a large scientific corpus of papers, reference material, knowledge bases and many other sources. We outperform existing models on a range of scientific tasks."
the paper: https://galactica.org/static/paper.pdf
the website to try the model: https://galactica.org/