Trump and Musk's plan for limitless artificial intelligence

Trump and Musk's plan for limitless artificial intelligence
Trump and Musk's plan for limitless artificial intelligence

Of the IAIA unrestrained to accelerate their development and also prevent them from becoming “ woke “. This is the wish of future American President Trump. No doubt under the influence of his campaign contributor Elon MuskElon Muskhe promised to repeal existing AI regulations put in place by current President Joe Biden. Provisions adopted in October 2023 and which, in any case, already had a fairly limited scope on the brakes placed on AI in mattermatter security.

For Trump, this decree is considered “dangerous” and hindering innovation in AI. But that's not all, since he sees in this decree a political act by the Democrats which ” imposes radical left-wing ideas on the development of this technology “. He wants an AI “ rooted in freedom of expression and human flourishing “. Words that seem to come straight from the mouth of Elon Musk who has advocated this philosophy since his purchase of TwitterTwitter and its transformation into X.

Two provisions of Biden's decree are in the crosshairs of Republicans and certain high-tech companies pushing for the unbridled development of AI. The latter criticize a provision which requires the assessment of the risks of large AI models. Companies in the sector must share their results from a test called “ red-team » with the government. The evaluation porteporte particularly on the vulnerability of AI to hacking.

Did you know?

The practice of “ red teaming » has its origins in Sun Tzu's military strategy in The of War. It consists of looking for hidden vulnerabilities that the adversary could exploit to plug them. In the case of AIs, it is about maintaining control over the ethically questionable initiatives that they could generate on their own.

Elon Musk speaks in the president's ear

This process of red teaming slows the pace of AI development and forces companies to disclose their design secrets, according to its critics. On the Republican side, it is rather another point that is being attacked: that which requires guaranteeing that AI models are free from bias likely to discriminate against certain groups based, for example, on race or gender. Too much ” woke » for Trump's supporters and first and foremost Elon Musk.

The latter especially has a lot to gain from a repeal of this decree. This would allow him to gain time from the delay he has fallen on his own AIs. Paradoxically, when opening ChatGPTChatGPT towards the general public, he called for the establishment of a moratorium on the development of AI to prevent them from becoming ethically out of control. It was, again, a pretext to try to catch up with OpenAI.

While waiting for the arrival of a repeal or a new decree, this announcement is cause for concern among AI specialists, who, like one of its sponsors, Yann LeCun, consider that it will soon be sufficiently intelligent to compete with humans and become dangerous for humanity.

-

-

PREV Alexianne has to pay 10 euros for her parking space while she paid 3 euros the day before, Charleroi airport explains
NEXT Tensions increase between Stellantis and the government of this country