Saturday, May 4, 2024

Vogon Today

Selected News from the Galaxy

Economic Scenarios

Was Sam Altman fired from OpenAI because AI became too dangerous? The Q* project

5 months ago babelfish

Now that Sam Altman is back, a mystery remains: why was he really fired on Friday 17 November? Was it just an inauspicious day? Has the Board of Directors gone mad? This sudden and risky decision could also revolve around a Reuters news story that suggests a connection to a discovery in the field of artificial general intelligence (AGI), which could threaten humanity.

In the days before Altman was sent into exile, several staff researchers wrote a letter to the board about a significant breakthrough – called Q* and pronounced Q-Star – that allowed the AI ​​model to “outperform humans on the most cost-effective tasks.” important."

Reuters sources said the AI ​​milestone was one of the significant factors that led to Altman's sudden dismissal by the board last Friday. T he concern was linked to the commercialization of an advanced model of artificial intelligence without understanding the socio-economic consequences which would be potentially enormous.

Furthermore, before Altman was fired, he could have referred to Q* at the Asia-Pacific Economic Cooperation in San Francisco: “Four times in the history of OpenAI—the last time was just in the last two weeks—I have managed to be in the room, when we kind of push back the veil of ignorance and the frontier of discovery forward, ” Altman said

This step forward would have triggered an internal conflict within OpenAI, which has its roots in an ideological battle between those who push for rapid progress in AI and those who support a slower, more responsible approach to development.

Reuters spoke to an OpenAI spokesperson who confirmed the existence of the Q* project and the letter to the board before Altman's firing.

So why is Q* a game changer? As tech blog 9to5Mac explains:

Currently, if you ask ChatGPT to solve a math problem, it will still use its text-on-steroids predictive approach to compile an answer using a huge text database and decide word-for-word how a human would respond. This means that he may or may not give the right answer, but in any case he has no mathematical competence, he will simply look for the answer somewhere.

OpenAI seems to have made a step forward in this area, successfully allowing an artificial intelligence model to actually solve mathematical problems never seen before and therefore not stored in some database. This development is known as Q*. Unfortunately the team didn't use a clever enough naming pattern to avoid something that looks like a pointer to a footnote, so I'll use the Q-Star version.

Q-Star's current mathematical ability is said to be that of an elementary school student, but this ability is expected to improve rapidly.

This technological development could be one of the first signs of the imminent commercialization of AGI, a form of artificial intelligence capable of surpassing humans.

AGI has the potential to surpass humans in every field, including creativity, problem solving, decision making, language understanding, etc., raising concerns about massive job displacement. A recent Goldman report highlights that 300 million layoffs could occur in the Western world due to artificial intelligence.

The Q* breakthrough and the rapid advancement of this technology now explain why the board suddenly fired Altman for his rush to develop this technology without studying the model's impact on how it threatens humanity.

Altman recently said, “ I think this is definitely the biggest update for people so far. And maybe the biggest one we'll have because from here on out, people will accept that powerful AI is, will happen, and there will be incremental upgrades… there was like the year the first iPhone came out, and then there there were all the subsequent ones .”


Meanwhile, Musk, who has warned about artificial intelligence threatening humanity, called the Reuters report “extremely concerning.”

Telegram
Thanks to our Telegram channel you can stay updated on the publication of new Economic Scenarios articles.

⇒ Sign up now ⇐


Minds

The article Was Sam Altman fired from OpenAI because AI has become too dangerous? The Q* project comes from Economic Scenarios .

This is a machine translation of a post published on Scenari Economici at the URL https://scenarieconomici.it/sam-altman-era-stato-licenziato-perche-la-ai-e-diventata-troppo-pericolosa/ on Fri, 24 Nov 2023 10:00:42 +0000.