Tech entrepreneur Elon Musk
Which: via REUTERS
Elon Musk apparently wants to build his own AI company. The Twitter boss recently called for a break in the development of artificial intelligence. The fact that he wants to compete with ChatGPT of all things is not without a certain irony.
EIt was only a few days ago that Twitter boss Elon Musk called for a broad-based break in artificial intelligence (AI). The development of AI applications should be stopped for six months, according to an open letter from researchers and entrepreneurs, which the eccentric multi-billionaire had co-signed. The reason: The rapid – and in the eyes of many frightening – progress of the chatbot ChatGPT. “AI stresses me out,” Musk recently explained.
That seems to be a thing of the past. Musk apparently has his own plans to found a new artificial intelligence start-up to compete with ChatGPT developer OpenAI. This is now reported by the “Financial Times”. Accordingly, the entrepreneur is currently putting together a team of researchers and engineers in the field of artificial intelligence. Among other things, the renowned AI expert Igor Babuschkin, who was most recently employed by Google’s parent company Alphabet in the developer company Deepmind, is named. Babushkin is an expert for those so-called large language models, as is ChatGPT.
In addition, talks are said to have taken place with a number of investors from aerospace company SpaceX and electric car manufacturer Tesla, i.e. the financiers of Musk’s other companies besides Twitter. They could also put money into his new company, so the hope.
Musk’s first ideas are said to have met with enthusiasm, the paper quotes an insider as saying. And the technical preparations should also be in full swing. Musk has already secured thousands of high-performance GPU processors from manufacturer Nvidia for the new project, reports the “Financial Times” with reference to confidants. The chips are essential when building AI systems, for example for text and image recognition and processing.
OpenAI now profitable start-up
ChatGPT’s recent successes may hurt Musk in another way. After all, it was he who co-founded its developer company OpenAI in 2015. Three years later, however, Musk left the board due to conflicts with management, allegedly over their lax attitudes towards AI security.
Musk has been warning of the dangers of artificial intelligence for years, and even regularly spoke of an existential threat to humanity. At that time, the goal of OpenAI was still “to advance digital intelligence in a way that benefits humanity as a whole, without being constrained by the need for financial return”. In the meantime, the AI forge has grown into a profitable start-up in which Microsoft recently invested several billion dollars.
Musk is by no means the only one who doesn’t want to lose touch in the race for the best AI models. In the past few days, a number of companies have announced that they will develop their own systems. The Chinese multinational Alibaba only presented its answer to ChatGPT on Tuesday. The bot called “Tongyi Qianwen”, which translates as “truth from a thousand questions”, is to be introduced on all Alibaba platforms – from online trading to card services.
“These are a handful of companies and they have it in their hands”
The technical possibilities are developing faster and faster. An artificial intelligence like ChatGPT is already being used by students and should also be included in curricula. “If you use that as a tool, it can be useful,” Prof. Alena Buyx, Chair of the German Ethics Council.
And AWS, the cloud subsidiary of US giant Amazon, announced two of its own models on Thursday – one is intended to generate text that the other can support in web searches. However, there was no dedicated chatbot among them.
Critics are now questioning whether Musk was ever serious about the required AI break. According to the Financial Times report, the multi-billionaire is said to have started recruiting high-ranking engineers for his project in January – well before the call for a break in development.
OK, AI?! is the WELT podcast for future optimists. Every Tuesday from 6 a.m., our AI experts talk to prominent guests from tech, politics, science and business. ThoseJoin the podcast among others Spotify, Apple Podcasts, Deezer, Amazon Music or directly by RSS-Feed.