Elon Musk will use Twitter to develop his artificial intelligence project

Twitter is a world of concepts and terms which Elon Musk You know you can take advantage of it. The owner of the social network said that it will use this data to store language models artificial intellectas ChatGPT Or the one he intended to create.

“In my opinion, Elon Musk is most likely using Twitter data, as well as reinforcement learning with human feedback and Tesla’s Dojo supercomputer to train language models on BasedAI,” said a user of the platform, to which the billionaire directly replied that “obviously. .”

A short answer, but in the context of the latest spin of the businessman and the information recently published, in which he forms a working team to create a language model that competes with ChatGPT and answers all the criticisms that have been made. Made on this platform.

You may also be interested in: Step by Step Add Subtitles to Twitter Videos

without forgetting that Twitter It is a social network in which text is the key to content, and there is a wide variety of terms and contexts that will greatly help to build a database that feeds learning. artificial intellect.

Which also contrasts with the high level of toxicity in the application. The Center for Combating Digital Hate (CCDH) conducted a study which found that hate speech on the platform has increased since it arrived Musk As the company’s CEO, for example, he went from 1,282 racist tweets against black people to 3,876 after Tesla was bought out.

A panorama that can also be heavily biased by an AI that has no conscience but relies solely on the information it provides.

You may also be interested in: Potentially Disruptive Scenarios Amid Advances in Artificial Intelligence

BasedAI, the key to this plan

In light of the artificial intelligence trend in recent weeks Elon Musk He mentioned the term several times: BasedAI. Although there is no concrete explanation yet, what is known about it is very much in line with what the businessman thinks about this technology, and some speculate that it will be the path or name of his project to compete with ChatGPT.

This concept defines an AI that is open and not biased, two points that the CEO of Twitter criticized for replicability, because these tools can be trained to track certain speeches, which is called WokeAI, or that they are closed to the organization. , refers to it as ClosedAI.

For this reason, the Twitter user mentioned this term in Musk’s comment, as his project will be based on the creation of an alternative to the training model, the details of which are not yet known, and only that it is known that recruitment is underway. Staff, including a researcher who recently left Google’s AI division.

You may also be interested in: Potentially Disruptive Scenarios Amid Advances in Artificial Intelligence

Musk vs. ChatGPT

After the chatbot had to create texts following a growing trend, a Tesla manager criticized its operation and regulation via a Twitter account.

Some of his comments were directed against Sam Altman, director of OpenIA, the company responsible for artificial intelligence, who claimed that “ChatGPT It’s awfully good. We are not far from dangerously powerful artificial intelligence. “

He then again emphasized the risks of progressing these technologies. “The dangers of awakening AI, in other words, lies, are deadly,” he wrote on his profile.

Criticism has extended to OpenIA, a company he helped found and which he claims has changed its vision after settling with larger companies.

“OpenAI was started as a non-profit open source company (that’s why I called it ‘open’ AI) as a counterweight to Google, but has now become a closed, top-grossing company effectively controlled by Microsoft.” This is not at all what I planned,” said the billionaire.