Episode 2534: Why Generative AI is a Technological Dead End
Something doesn’t smell right about generative AI. Earlier this week, we had a featuring a former Google researcher who described large language models (LLMs) as a “con”. Then, of course, there’s OpenAI CEO Sam Altman who critics both inside and outside OpenAI, see as a little more than a persuasive conman. Scam or not, the biggest technical problem with LLMs, according to Peter Vos, who invented the term Artificial General Intelligence (AGI), is that it lacks memory and thus are inherently incapable of incremental learning. Voss, the current CEO of Aigo.ai, argues that LLMs therefore represent a technological “dead end” for AI. The industry, Voss argues, has gone “fundamentally wrong” with generative AI. It’s a classic economic mania, he says. And, as with all bubbles in the past - like Dutch tulips, internet DotComs or Japanese real-estate - it will eventually burst with devastating consequences.
Peter Voss is a pioneer in AI who coined the term 'AGI' (Artificial General Intelligence) in 2001. As an engineer, scientist, inventor, and serial entrepreneur, he developed a comprehensive ERP package, growing his company from zero to a 400-person IPO in seven years. For the past 20 years, he has studied intelligence and AI, leading to the creation of the Aigo engine, a proto-AGI natural language intelligence that is revolutionizing call centers and advancing towards human-level intelligence.
Named as one of the "100 most connected men" by GQ magazine, Andrew Keen is amongst the world's best known broadcasters and commentators. In addition to presenting the daily KEEN ON show, he is the host of the long-running How To Fix Democracy interview series. He is also the author of four prescient books about digital technology: CULT OF THE AMATEUR, DIGITAL VERTIGO, THE INTERNET IS NOT THE ANSWER and HOW TO FIX THE FUTURE. Andrew lives in San Francisco, is married to Cassandra Knight, Google's VP of Litigation & Discovery, and has two grown children.
Keen On America is a reader-supported publication. To receive new posts and support my work, consider becoming a free or paid subscriber.
This is a public episode. If you'd like to discuss this with other subscribers or get access to bonus episodes, visit keenon.substack.com/subscribe
Episode 2300: Sandra Matz makes the Case for a Data-Driven Science of Predicting and Changing Human Behavior
Episode 2299: Jill Kastner explains why everything old is new again in international politics
Episode 2298: Adam Chandler on the fatal contradiction at the heart of American capitalism
Episode 2297: Louis Ferrante on why the Mafia Killed JFK
Episode 2296: Adi Jaffe on how to free yourself from addiction forever
Episode 2295: Paula Whyman on how to save the American environment - one wild mountaintop at a time
Episode 2294: Larry Downes' non-MAGA plan to shrink the Federal bureaucracy
Episode 2293: David Masciotra on why Kamala Harris should have gone on the Joe Rogan show
Episode 2292: Chris Schroeder on how America now swims in an ocean of black swans
Episode 2291: Michael Scott-Baumann on the hopelessness of the Palestinian situation
Episode 2290: Marshall Poe on why 2024 was a bad year for most podcasters
Episode 2289: Gary Marcus on how Artificial General Intelligence (AGI) is, in the long run, inevitable
Episode 2288: Simon Kuper on the chilling parallels between MAGA America and Apartheid South Africa