The existential risk from AI – and from people misusing it | synthetic intelligence (AI)

Relating to Jonathan Freedland’s article about AI (The future of AI is bright – humans must work together to face the threat to civilization, May 26), is not worrying about whether or not an AI is “emotional” slightly like worrying about whether or not a synthetic limb is “alive”? There’s nonetheless no proof that “absorption” is a factor. Very like life, it is a bunch of various abilities, and “AI” (i.e. finite synthetic intelligence) is not prone to reproduce greater than two of these abilities.

It’s because it’s an try to breed the perform of solely a small a part of the human mind: particularly, the evolutionarily new half. Our motivation to pursue self-interest comes from a billion-year evolution of the traditional mind, which isn’t based mostly on AI. The true hazard is from people misusing AI for their very own functions, and the truth that the mechanisms we have constructed to acknowledge different organisms with minds like ours (as Friedland identified) are too simple. They’re fooled by superficial proof.
Roger Haynes
London

Isaac Asimov’s guide I, The Robotic makes helpful studying. I quote the introductory web page as follows.

Three Legal guidelines of Robotics:
“1) A robotic might not injure a human being, or, by way of inaction, permit a human being to be harmed.
2) A robotic should obey orders given by people besides the place such orders battle with the First Regulation.
3) A robotic should defend its existence so long as this safety doesn’t battle with the primary or second regulation.

Asimov’s predictions, made 70 years in the past, envisioned a time in 2058 when these legal guidelines could be mandatory. Issues turned out sooner than he anticipated.
Professor Paul Huxley
London

Your article (Yes, you should worry about AI – but the matrix analog hides a more dangerous threat, May 30) confirms a lot of my suspicion about AI being the “existential risk” we have now to worry. It feels a little bit like Y2K horror. It appears to comply with the identical playbook. First, set up a compelling future trigger for company and political concern, then promote the “answer,” which in fact will price severe cash in analysis and consulting charges.

Consolidation of damages Florian statements are clear, present and clear in view, and it isn’t troublesome to search out means to take care of them. We do not want snake oil salesmen for this.
Phil Hyde
Coventry