A few days before that Sam Altman was fired from OpenAIsome researchers communicated to the company’s board of directors a scientific discovery that – according to sources cited by Reuters in reporting this exclusive news – could constitute a danger to humanity.
Their letter would have triggered a “coup d’etat” by the board, which at the time consisted of Altman and five other people.
The OpenAI case OpenAI, who are the three people on the board who had the fate of the company in their hands by Pier Luigi Pisa 21 November 2023
At least four – the scientific director of OpenAI Ilya Sutskeverthe advocate of Effective Altruism Helen Tonerthe robotics expert Tasha McCauley and the entrepreneur Adam D’Angelo – voted to urgently remove the company’s CEO, concerned that Altman might commercialize a powerful technology without adequately assessing its risks.
The case To understand what happened to Sam Altman, you need to know what OpenAI is really like by Pier Luigi Pisa 19 November 2023
In the letter from the OpenAI researchers – which Reuters has not yet had the opportunity to read – reference is made to a new company projectcalled Q* (read Q Star), which would concern a type of AI very close to breaking latest newsgeneral artificial intelligence that is capable of replicating – and in some cases surpassing – human capabilities.
OpenAI, founded in San Francisco in 2015 by a group of researchers and entrepreneurs, including Sam Altman and Elon Muskhas become world famous for its generative artificial intelligence capable of writing (ChatGpt) and create images (Give her) as a human would. But his real mission, from the beginning, is to achieve breaking latest news “for the good of humanity”.
OpenAI defines breaking latest news as an autonomous system that surpasses humans “in most activities that have an economic value”.
While generative AI relies on predefined models, and imitates human creativity without truly understanding the meaning of what it creates, breaking latest news is a type of artificial intelligence much more advanced.
breaking latest news can learn and understand specific problems like a human would. Furthermore, she can adapt to new situations without having to be trained specifically for each task.
Some OpenAI researchers believe that Q* is close to the capabilities of breaking latest news because it has been able to solve some mathematical problems. They would be rather “simple” problems, the kind that are faced to elementary schoolsbut for OpenAI this result would be of enormous importance.
We know, in fact, that ChatGpt is very good at writing texts – whose answers vary each time precisely because they are the result of a statistical model – while committing numerous errors in mathematics, where only one correct answer is allowed. This ability of Q*, according to OpenAI researchers, would bring artificial intelligence closer to the reasoning capabilities of the human mind.
Last November 6th, on the occasion of DevDaySam Altman closed the day dedicated by OpenAI to developers by stating: “In 2024 what we launched today [le intelligenze artificiali personalizzate chiamate GPT, nda] it will seem antiquated compared to what we are building for you.” Evidently it wasn’t just a good marketing phrase.
Last November 16, while attending a conference in San Francisco, Altman himself cryptically referred to important new advances: “For the past two weeks, I have had the opportunity to be in the room as we pushed back the veil of ignorance and moved the frontier of discovery forward.”
The next day, Sam Altman was fired from the board.
The entrepreneur’s exile it lasted just four days. Pressure from OpenAI employees and investors – Microsoft above all – brought Altman back to the company. Sam is the CEO of the company again. And his power is even greater now.