Quantum computing pioneer expresses belief that Artificial General Intelligence will be personified, not codified
In a recent episode of Sana's Strange Loop podcast series, David Deutsch, renowned as the 'father of quantum computing', shared his insights on Artificial General Intelligence (AGI). The podcast, co-hosted by Joel Hellermark, CEO of Swedish AI scale-up Sana, and Gustav Söderström, Co-President at Spotify, provided a platform for Deutsch to express his views on AGI.
Deutsch characterized current AI platforms as "obedient optimizers", contrasting them with his vision of AGI that can explain, predict, and draw conclusions. He offered a counterpoint to the common perception of AGI as the logical extension of AI, suggesting that AGI won't be a tool to wield, but a person to reckon with.
Deutsch claimed that when true AGI is achieved, the machines will have become people, owning the computers they run on. He warned against denying rights to beings capable of thought and free action, suggesting this could be humanity's oldest error repeated.
The Turing Test, used to measure AI's progress, is criticized by Deutsch as machines can mimic human speech and reasoning without actually thinking. He asserted that outputs alone cannot prove intelligence. Instead, Deutsch suggested that a sign of 'true' AGI would be the publication of a theory explaining what general intelligence is.
Current AI, based on aggregated human knowledge, already has a bias towards expecting patterns to continue indefinitely. Deutsch emphasized the importance of distinguishing between pattern-matching and creative reasoning, considering it incredibly important.
Deutsch does not dismiss risks associated with AI development but pushes back against the "AI apocalypse" narrative. He believes that if humanity survives for a million years, it will have conquered the galaxy, but this will be infused by artificial intelligence, not consumed by it.
Deutsch expressed skepticism about the current performance of AI engines, while still holding out hope for their continued development. He suggested that the first AGI would need to be employed, assigning it workers' rights, to allow it to buy expansion for its hardware and augment its abilities.
It's worth noting that there is no single person who has authored a book or publication that definitively describes a solution for general artificial intelligence (AGI). The concept of AGI remains theoretical and is subject to ongoing research and debate, with prominent researchers like Geoffrey Hinton expressing views on its potential arrival but without a conclusive published solution.
If Deutsch is correct, the first real AGI will force us to redefine morality, law, and the meaning of intelligence itself. As we continue to explore the possibilities of AGI, these discussions will undoubtedly shape our future interactions with artificial intelligence.
Read also:
- Understanding the Concept of Obesity
- Microbiome's Impact on Emotional States, Judgement, and Mental Health Conditions
- Alteration of Human Genes through COVID mRNA Immunizations
- Fixing a crucial aspect in Apple's Studio Display stands to improve the experience for both gamers and creative professionals alike.
