The Californian company OpenAI, behind the ChatGPT phenomenon, on Tuesday launched GPT-4, a new version of generative artificial intelligence technology that operates the famous chatbot and is one step closer to computer programs as "intelligent" as humans.

Microsoft, which has invested billions of dollars in the start-up, announced in the wake of having integrated GPT-4 into Bing, its search engine already equipped with ChatGPT features for a month.

"GPT-4 is a large multimedia model, less skilled than humans in many real-life scenarios, but as good as humans in many professional and academic contexts," OpenAI said in a statement. "For example, he passes the exam to become a lawyer with a score as good as the top 10%. The previous version, GPT 3.5, was at the level of the worst 10%," she said.

Enthusiasm and controversy

ChatGPT arouses a lot of enthusiasm, but also controversy, since it is open access and used by millions of people around the world to write essays, lines of code, advertisements or simply to test its abilities.

OpenAI has thus established itself as the leader in generative artificial intelligence (AI) with its programs producing texts or, like DALL-E, images. Its boss, Sam Altman, recently explained that he is now working towards so-called "general" artificial intelligence, that is to say programs with human cognitive abilities.

"Our mission is to ensure that general AI -- AI systems that are smarter than humans in general -- benefits all of humanity," he said on the company's blog on Feb. 24.

ChatGPT at the University

The multimedia capabilities of GPT-4 are a step in this direction. Unlike previous versions, the new model is indeed equipped with vision: it can process text but also images. Instead, it only generates text. It will be available on ChatGPT, but without the ability to provide it with images for now.

"The power of the algorithm will increase, but this is not a second revolution," commented Robert Vesoul, CEO of the French company Illuin Technology. "We didn't go from the Moon to Mars."

Jim Fan, an AI specialist who worked for Google and OpenAI, and now at Nvidia, was more impressed. "GPT-4 can now apply to study at Stanford (a prestigious American university). His ability to reason is NEVER SEEN! " he tweeted, admitting to having received worse results in some exams than the model.

The AI rush

The ChatGPT craze has launched a race for the most successful and useful generative AI among tech companies. Microsoft has added tools based on OpenAI's language models on Bing but also on other platforms, from office software to its browser, Edge. Google has launched Bard, a competitor to ChatGPT.

Google also unveiled new features Tuesday that will allow users of Gmail, Google Docs and Google Cloud, among others, to generate emails, ad campaigns and other documents. The social network Snapchat and e-commerce applications have added chatbots of this new generation.

Blunders

These rapid deployments have resulted in many hiccups, from an advertisement for Bard with factual errors to Bing declaring its love for a New York Times reporter. "Despite its capabilities, GPT-4 has similar limitations to previous models," OpenAI pointed out. "He is not yet totally reliable (he hallucinates, invents things and makes mistakes of reasoning)."

The company announced that it had hired more than 50 experts to assess the new dangers that could emerge, for cybersecurity for example, in addition to the risks already known (generation of dangerous advice, defective computer code, false information, etc.). Their feedback and analysis should make it possible to improve the model. "In particular, we collected additional data to ensure that GPT-4 refuses user requests about the manufacture of hazardous chemicals."

In the immediate future, generative AI worries many intellectual and creative professions, who already imagine themselves reduced to the role of managing chatbots to extract the best texts and images. "What will be left for humans?" tweeted Elon Musk, one of the entrepreneurs who fear that artificial intelligence could one day take over humans.

  • Tech
  • OpenAI
  • Bing
  • Artificial Intelligence (AI)