Kilpajuoksua kohti AGIa on mahdotonta hillitä. Ei ihme, sillä AGI on tärkein tapahtuma ihmiskunnan historiassa. AGI on tekoäly, jolla on ihmisen kaltainen yleinen älykkyys.
Mark Zuckerberg haluaa pikaisesti luoda AGIn, jonka kanssa voisi jutella erityisillä silmälaseilla. Silmälaseissa on ääni ja kuvayhteys AGIin.
Zuckerbergin Meta käyttää tänä vuonna 10 miljardia dollaria eli lähes yhden marinin (1 Marin = 10 miljardia euroa) Nvidian suorittimiin. Vuoden 2024 lopulla H100 -suorittimia on 600.000 kpl.
Vertailun vuoksi OpenAI:n ChatGPT:n opettamisessa käytettiin ehkä 25.000 kpl A100-suoritinta. H100 on opetuksessa yhdeksän kertaa nopeampi kuin A100 ja tuotantokäytössä 30 kertaa nopeampi. Metalla on siten 216 kertaa enemmän laskentatehoa opetuksessa ja 720 kertaa enemmän tuotantokäytössä.
Meta on jo alkanut opettamaan Llama3-kielimallia tavoitteena AGI. Llama3 julkaistaan avoimena koodina eli lähdekoodi tullaan julkaisemaan.
Ilmeisesti Zuckerberg ei piittaa varoituksista, että AGI voi olla peruuttamaton askel kohti ihmiskunnan nopeaa loppua.
Muutkin kuin Meta haluavat AGIn. Esim. Microsoft on jo käyttänyt 13 miljardia dollaria OpenAI:n ChatGPT:en. OpenAI ei kuitenkaan aio antaa MIcrosoftille yksinoikeutta tulossa olevaan AGIin.
- zuckerberg_poker.jpeg (88.16 KiB) Katsottu 264 kertaa
"Achieving AGI may also feel like a point of no return for the human race — with Google CEO Sundar Pichai saying as far back as 2018 that the field of AI research is "more profound than electricity or fire." Last year, dozens of experts and prominent figures — including OpenAI CEO Sam Altman and Microsoft founder Bill Gates — signed a statement stressing the collective need for humanity to mitigate "the risk of extinction from AI" alongside other societal-scale risks such as pandemics and nuclear war. That said, many scientists think humanity can never build AGI.
Mark Zuckerberg said Meta will have "an absolutely massive amount of infrastructure" in place by the end of the year to prime it for training an artificial general intelligence model.
But Zuckerberg announced in an Instagram reel that the company is buying 350,000 Nvidia H100 graphics processing units (GPUs) — some of the most powerful graphics cards in the world — which are key to training today's best AI models.
This will more than double Meta's total computing power for AI training, with Meta aiming to wield computing power equivalent to 600,000 H100 GPUs in total.
Nvidia's H100 is the newer version of the A100 graphics cards, which OpenAI used to train ChatGPT. Our best available knowledge, based on unverified leaks, suggests OpenAI used roughly 25,000 Nvidia A100 GPUs for the chatbot's training — although other estimates suggest this number is lower."
Artificial general intelligence — when AI becomes more capable than humans — is just moments away, Meta's Mark Zuckerberg declares
https://www.livescience.com/technology/ ... g-declares
Livescience 20.1.24
The H100 GPU is up to nine times faster for AI training and thirty times faster for inference than the A100. The NVIDIA H100 80GB SXM5 is two times faster than the NVIDIA A100 80GB SXM4 when running FlashAttention-2 training.
https://www.cudocompute.com/blog/compar ... h100-gpus/
"Meta CEO, Mark Zuckerberg, is saying that the LLAMA 3 large language model will be providing open source AGI.
Zuckerberg feels that people will wear glasses to hold their connection to AGI. Glasses will provide the video and audio connection from people to AGI."
Meta Launching LLAMA 3 and Opensource Eventual AGI
https://www.nextbigfuture.com/2024/01/m ... e-agi.html
NextBIGFUTURE 19.1.24
"OpenAI was co-founded by Sam Altman, Elon Musk, Ilya Sutskever, and others in 2015, with a non-profit model.
However, in 2019, the company set up a subsidiary with a for-profit model, capped at 100 times the investment. Microsoft invested $1 billion in the company to begin with – it has since increased that investment to $13 billion so far.
While LeCun's remarks have been scathing, OpenAI and Altman have laid out their vision for AGI, both over the long and short terms. Altman and Musk also agree on at least one thing – AGI will be the "most important event" in world's history."
Meta's Chief AI Scientist Calls OpenAI A 'Contract Research House' For Microsoft, Says They Won't Be The First To Develop AGI
https://www.alsahm.com/news/content/met ... 2023-12-28
Benzinga News 28.12.2023
"Microsoft is a major investor in OpenAI and, according to rumors, wants to secure up to 49 percent. However, the software giant is not to get exclusive rights to ChatGPT or other models.
Altman also confirmed that OpenAI is working on a video model. However, he said it is too early to talk about a release date.
A conversation about AI with one of the industry's big players inevitably leads to the question of how to safely and responsibly use AI.
Google just released a paper on the topic and high-ranking representatives from Alphabet, Google and Deepmind issued a detailed statement on AI progress as part of the release."
Microsoft not to get exclusive rights to OpenAIs models
https://the-decoder.com/microsoft-not-t ... is-models/
The decoder, 18.1.24
SDP pitäisi kieltää hyvien tapojen vastaisena.
