"The AI may in fact be conscious". Wow, this topic moves FAST! Anthropic's latest Terms of Reference forbid 'cruel behavior towards Claude"

Three weekends ago the news was a random youngster from one Frontier Lab going on every TV show to say that there's a 10% chance of AI wiping out humanity, albeit it with no explanation of the "how" (or in fact the "why").

This led to the late September debates around what people meant by "slow it down" while they pushed harder on the gas pedal.

Then there was a brief break in the existential debate to allow time for us to rename AI as #SuperIntelligence.

It's heated up again now with this weekend's debate being about whether AI is conscious and why this discussion is even coming to pass. This topic has been on the boil for much of year (and far longer in niche circles) having spike initially in Feb/March then calmed and is become headline again now. The Friday 9th "All-In" podcast gave a very articulate view from the four presenters (and minimal politics for those who find that stops them listening). Equally insightful is this post by Henry Shevlin which is well worth reading (though he mentions a refined version may arrive once he's got of the plane) - https://lnkd.in/eStXkqTi

Now, to most people reading this in the UK this whole post will seem utterly nuts. Two reasons - one we haven't drunk enough of the kool-aid yet so while mention of AI, sorry #SI, maybe as omnipresent as over on the US West Coast, the view of the impact is orders of magnitude less. Most people are still at the "it can find a file on the network I thought I'd lost" or "it has fixed the background on the photo for my passport application" level.

Secondly, even mismatched hype-levels aside, the actual "news" is taking a week or two to get it's way over into the UK professional mainstream. Most people are off Twitter/X these days and there are quite enough nearer-term crisises in the UK and Europe with tax rises, bond prices and increasing fear about a grey-zone hybrid war. The hallucinations/ aspirations (*delete as applicable) of tech-bro billionaires hasn't quite registered in most people's minds despite how profound and near-term it potentially is.

Anyhow, enough of this as you're far better off ready what Henry has written, or diving into this recent post by Allegra covering off some of the potential drivers for the Frontier Labs' behaviours: "Where Ethics and Caution meet" - https://lnkd.in/eSGU2NBw

One would hope that this thinking by Henry and Allegra and many others would enrich the debate but it feels that this is getting politicised again. Let's see how the coming days pan out and whether the US focus on their mid-term elections and gas prices distracts enough that it calms down.

***Follow the InnoEthics initiative for ongoing insights on these topics and learn about critical concepts such as Sociopathic Intelligence and a-ethical AI

(Image by Nat on Unsplash)

By Paolo Cuomo

Keywords: AI, AI Ethics, AI Safety

Follow Us On