Skype cofounder Jaan Tallinn believes the latest surge in the development of artificial intelligence (AI) technologies is “different” to previous advances in the controversial subject area.

The Estonian entrepreneur, who is working closely with a number of AI companies and researchers through the Future of Life Istitute (FLI) that was backed by Elon Musk and Stephen Hawking earlier this year, told Techworld that the history of AI development has traditionally been a series of summers (peaks) and winters (troughs).

Honda's humanoid robot Asimo of how humans are making increasingly intelligent robots ©Flickr/Ars Technica
Honda's humanoid robot Asimo of how humans are making increasingly intelligent robots ©Flickr/Ars Technica

In the summers, AI companies make advances and get lots of investment but in the winters the money dries up and companies rename it to something like “machine learning”, Tallinn said. 

“There’s something very different about this AI summer,” he said at a hotel in London’s Soho district last week, while plugging his new instant messenger app, Fleep, which he has cofounded with ex-Skype colleague Henn Ruukel. “There are quick ways to make money from marginal advances in AI. Once you make a ranking algorithm one percent better, that immediately means a few hundred million dollars for Google.”

The claim comes amid growing fears that machines are going to surpass the capabilities of humans in both the jobs market and many other areas of life. 

Through Future of Life, which was cofounded with MIT professor Max Tegmark, Tallinn said he wants to bring together two schools of thought: the “physicists and the philosophers” who are concerned about the long term about what happens when humans “lose control of atoms” and, on the other side, the technology companies who are focused on the short term and making machines more competent at helping us.  

Jaan Tallinn at the Ham Yard Hotel in London's Soho neighbourhood ©Techworld/Sam Shead

Tallinn has invested in three AI companies, including the UK's DeepMind, which was acquired by Google last year for a sum thought to be around $400 million (£264 million). The other two, Vicarious and Dextro, are in New Jersey and New York respectively.

Despite investing in companies working on advancing AI, Tallinn said he tries not to lean one way or the other.

“By doing that you get that us vs. them effect,” he said. “I try to be the mutual guy in the middle that seems reasonable to both sides.”

More precautions needed

An open letter published by the Future of Life Institute in January warned that more precautions need to be taken around the further development of AI. 

The letter argues that scientists and technologists need to safely and carefully coordinate and communicate advancements in AI to ensure it does not grow beyond humanity's control.

“Because of the great potential of AI, it is important to research how to reap its benefits while avoiding potential pitfalls,” reads the letter. “Our AI systems must do what we want them to do.” 

Musk, the cofounder of SpaceX and Tesla, a member of the FLI’s scientific advisory board alongside actor Morgan Freeman and world-renowned Cambridge University professor, Stephen Hawking, has previously said that uncontrolled development of AI could be "potentially more dangerous than nukes". 

Experts at some of the world's biggest tech corporations - including IBM's Watson supercomputer team, Google, Microsoft Research and Amazon -  also signed the letter. 

Other signatories include the entrepreneurs behind artificial intelligence companies that Tallinn has backed, such as DeepMind and Vicarious.

Musk has pledged $10 million (£6.6 million) to help fund research into the development of AI through the FLI. 

Tallinn said there has been a "good response" from the academic community and the FLI is now "sifting through" them and deciding which ones to fund. 

Many technology companies are racing ahead with their own AI research in a bid to cash in on the technology's potential but Tallinn doesn't think any rules or regulations should be introduced just yet. 

"I think it’s too early to think about very concrete monitoring mechanisms," said Tallinn. "I think it’s more important right now to build concensus in the industry and academia around what are the things that would have a chilling effect."