Illustration of a computer with quotes, binary numbers, and abstract shapes.

Illustration: Shoshana Gordon/Axios

Forget cryptocurrency and blockchain: the technology conversation at this year’s World Economic Forum in Davos is all about the rise of artificial intelligence, particularly the ChatGPT text generator.

Why it matters: Tools like OpenAI’s ChatGPT and imagers like Stable Diffusion and Dall-E have been in the works for years, but even techies in the Davos crowd are surprised at how quickly they’ve matured.

What they are saying: In panels and side conversations at this week’s meeting in Davos, Switzerland, as well as last week’s DLD Conference in Munich, everyone wants to talk about this latest generation of generative AI tools, from how they’re personally experimenting with it to how to see it reshaping their businesses and lives.

  • The CEO of a major tech company I spoke to on the sidelines told me he knew all about the wide language model approach underlying these generative AI tools, but he wouldn’t have predicted even six months ago that they would emerged as the game. -changers that are outlined to be.

The panorama: It’s clear that generative AI has captured the public imagination like no technology has since the arrival of the iPhone in 2007.

  • Everyone is still trying to understand how these technologies will change the way they live and work, with some incredibly excited, some fearful, and many just busy writing ChatGPT queries.

Optimists see a world in which AI gives knowledge workers superpowers and accelerates the time needed to achieve health and sustainability breakthroughs.

  • Hanzade Dogan, president of Turkish e-commerce company Hepsiburada, highlighted the opportunity AI has to dramatically reduce the costs of expensive services, expanding access to legal aid, healthcare and more.
  • “Or, if we’re wrong, it could be the dystopia of our world,” he said during a panel I moderated on Wednesday at the Forum. “That’s how serious what we’re dealing with is.”
  • Investor Jim Breyer has poured money into a dozen companies working to use AI in a variety of healthcare applications, including early detection of prostate and breast cancers.
  • “I think the biggest commercial application of AI will be in precision medicine. Tough stop,” Breyer said.

Yes, but: Concerns range from the inevitable flood of AI-generated misinformation to biases built into systems that have been trained on real-world data that is riddled with stereotypes and dominated by rich countries. CEO Tom Siebel He said it’s important to understand biases in the data when choosing which problems to target with AI.

  • For example, he said his company rejected the idea of ​​working on a large contract with the military to use AI to help determine Army promotions, noting that it would inevitably recommend white male West Point graduates.

  • “We are not going to touch it, and my recommendation is that you do not touch it either,” he said.

Brett Solomon, CEO of Access Now he told Axios earlier this week that he is concerned that this new crop of artificial intelligence technologies will be another weapon used against human rights activists, journalists and others.

  • “Given the fact that civil society is already under attack, our ability to defend against AI-generative phishing attacks, impersonations, and misrepresentations will put us at even greater risk,” Solomon said.

another big concern is what AI will mean for jobs.

  • The experts I spoke to agree that these changes are inevitable and that the best thing governments can do on this front is to help train workers for a reformed world. (I’m moderating another panel for the Forum on Friday focused specifically on AI and jobs.)

Whats Next: A big question is how regulators will approach the technology. The EU is already working on an AI Law, which is intended to be the first comprehensive legislation to govern such technology.

  • Another key debate is whether AI systems should essentially be able to display their work, or whether it’s good enough that they only improve their accuracy.

The bottom line: Everyone agrees that today’s generative AIs need big improvements, especially given their tendency to be sure wrong.

  • At the rate this field is advancing, those improvements may not take long.