Mastodawn

13h ago

@ChrisMayLA6 This may well be the worst timeline in the history of humanity for the rise of #AI. Indeed, will humanity survive the AI age?!

Emeritus Prof. Christopher May 13h ago

Well, we're back to the Terminator meme, I guess....

13h ago

@ChrisMayLA6 We jest but with the way things are (Mad King Trump and the power crazed billionaires) the Terminator scenario becomes ever more believable! 😱

Emeritus Prof. Christopher May 13h ago

Yes, although the tech-folk here on Mastodon still mobilise some good arguments for why its overly alarmist....

13h ago

@ChrisMayLA6 Yes, fear mongering is to be avoided but, and I may be wrong about this, caution would be wise. Maybe I should stop reading articles about AIs getting close to building themselves.

Peter 13h ago

There is no AI yet.

There is pattern recognition, which science and medicine make great use of. There is LLM- style text and image generation, which ordinary folk like to play with, and there are programmed robots which carry out planned actions.

There are robots which can get between two points, chosen by a human, autonomously, sometimes while carrying a box if required to by a human.

The gulf between all of these is, as yet, unbridged. They rely on completely different versions of "AI".

The problem is humans giving an "AI" system an instruction and letting it work out the best process to carry it out—without humans being in the loop prior to execution. THAT is dangerous.

We're at least a human generation away—at the soonest—from anything resembling artificial sentience, which is what most people think of when Hollywood mentions "AI".

Emeritus Prof. Christopher May 13h ago

@PeterLG @alex_p_roe

Which is also why I use the term 'AI & associated technologies'.... I'll be interested to see if we really are a generation away from AI sentience (25 years seems like a long time in technological development terms)

Peter 12h ago

My thought is that is will take quantum computing to provide the necessary bit crunching oomph to reach AS.

Real QC, not the current single- or multiple-qubit experimental stuff that takes a room full of gear to barely work. (Flashbacks to early IBM...😄)

(The lastest QC stuff IS exciting though.)

@PeterLG @ChrisMayLA6 Indeed, QC is beginning to make progress from what I’ve read recently. And it could lead to genuine AI or something that gives the impression as LLMs sometimes do, that they are intelligent.

Enola Knezevic 12h ago

@ChrisMayLA6 @PeterLG @alex_p_roe I am sure we are going to render the planet uninhabitable to humans before we manufacture anything that could be called intelligent, sentient, or conscious.

12h ago

@rhelune @ChrisMayLA6 @PeterLG Hmm, good point. It’s a bit of a race towards who knows what end - either hell on earth or paradise. Too soon to know!

Ken Milmore 12h ago

@ChrisMayLA6 @PeterLG @alex_p_roe I am reminded of Paul Merton on HIGNFY some years ago despairing that we aren't all flying around with jet packs like he'd once been promised. Sometimes we have a false expectation of what technology will bring us.

11h ago

@kbm0 @ChrisMayLA6 @PeterLG Am still waiting for flying cars! The sci-fi of around 100 years ago was a reasonable predictor of some of the tech which now exists so maybe we’ll see certain super technologies in the next 100 years or so, or less seeing as the cycle is continuous. Asimov’s robots are edging closer!

Peter 11h ago

Just as long as those flying cars use true sci-fi gravity nullifying tech, and not dirty great spinning blades of death like most companies are playing with at the moment! 😅

@kbm0 @ChrisMayLA6

@PeterLG @alex_p_roe @kbm0 @ChrisMayLA6

D. G. Marshall 11h ago

I remember a story in Popular Mechanics (IIRC), that had a mock news report stating there was now a no-fly order in force over the nudist beach, because distracted drivers kept crashing their flying cars.

Trebach 6h ago

@davidtheeviloverlord @PeterLG @alex_p_roe @kbm0 @ChrisMayLA6 It has happened before with boats: https://www.nbcnews.com/id/wbna4890829

Barge sinks after nude sighting in Texas

Two people were hospitalized with minor injuries Sunday after a rented double-decker barge sank near Hippie Hollow, Texas, when partygoers crowded one side of the vessel, hoping to catch a glimpse of nude sunbathers.

NBC News

6h ago

@trebach @davidtheeviloverlord @PeterLG @kbm0 @ChrisMayLA6 Ha! Rubber necking has always been responsible for accidents!

@PeterLG @alex_p_roe @kbm0 @ChrisMayLA6

@stevewfolds 11h ago

@PeterLG @alex_p_roe @kbm0 @ChrisMayLA6
https://www.gravityresearchfoundation.org/historic

@stevewfolds 10h ago

Gravity Research Foundation

The founding history of the Gravity Research Foundation

Gravity Research Foundation

@stevewfolds @PeterLG @kbm0 @ChrisMayLA6 I guess we may eventually figure out gravity and LLMs may well help us do this. If we do, we might be able to do all sorts of whizzy things. I’m not going to hold my breath though.

Peter 8h ago

@stevewfolds @kbm0 @ChrisMayLA6

Not LLMs; they have no ability to make intuitive leaps. That, I suspect, will take real intelligence—either human or artificial.

6h ago

@PeterLG @stevewfolds @kbm0 @ChrisMayLA6 No, LLMs can’t intuit a thing but they can be better at joining dots than us meat sacks - assuming the connections are not hallucinations, that is.

@PeterLG @kbm0 @ChrisMayLA6 Well, some #flying-cars are using lots of little enclosed blades which are a wee bit safer until one cluster stops working and drops a formerly flying car on people’s heads. One tech I’m watching is fluidic propulsion by a company called Jet Optera https://www.jetoptera.com/ - it looks quite promising even if progress appears to be slow.

Jetoptera | Revolutionary Propulsion System

aoanla 11h ago

@kbm0 @ChrisMayLA6 @PeterLG @alex_p_roe jetpacks, and flying cars, are a great example of possible tech that turned out to not be practical. You could have a jetpack now, but it turns out that carrying volatile fuel on your back and having a very hot exhaust stream perilously close to your legs is a bad idea.
"AI" may be similar - the negatives of LLMs probably outweigh the positives that you might have imagined from SF.

Ken Milmore 11h ago

@ChrisMayLA6 @PeterLG @alex_p_roe Philosophically, we should probably consider the flip side: Suppose an AI system is ever produced which is considered by general agreement to be sentient. There are dark precedents for putting sentient beings into forced labour. The way to take the wind out of this before it ever happens would be to legislate in advance against any monetisation or ownership of an artificial sentience. If there's no money in it, I can't see the tech bros holding their interest.

Emeritus Prof. Christopher May 11h ago

@kbm0 @PeterLG @alex_p_roe

There's a micro-trend in SF that explores what happens when sentient robots gain political rights.... interesting idea

Ken Milmore 11h ago

@ChrisMayLA6 @PeterLG @alex_p_roe If it was made illegal to own or monetise an artificial sentience then we might see the AGI hype pivot quite quickly I think. The emphasis would suddenly be on pushing AI as just another appliance and not really that intelligent at all.

@kbm0 @ChrisMayLA6 @PeterLG Good luck getting current bought and paid for governments to go for that especially in the US where someone seems more than happy to be compensated for whatever law someone else is prepared to pay for. AI is a global issue but the approach to its regulation is local and even then, piecemeal.

https://www.forbes.com/sites/anishasircar/2026/06/10/ai-owned-companies-argentina/

MarjorieR 11h ago

@ChrisMayLA6 @kbm0 @PeterLG @alex_p_roe it seens Thiel and Milei seems to be intent on turning this from science fiction into fact.

Argentina Wants To Let AI Own Companies. Here’s What That Means

Argentina is considering legislation to allow AI-run companies with no human owners or executives. Supporters see innovation; critics warn of accountability risks.

Forbes

Ken Milmore 11h ago

@marjolica @ChrisMayLA6 @PeterLG @alex_p_roe I wonder what's really behind this? Is it to do with limitation of liability? If you want to run an AI system but don't want to take any responsibility for the damage caused by its mistakes then I suspect this is the sort of legal framework you might choose.

@kbm0 @marjolica @ChrisMayLA6 @PeterLG With Theil and Milei involved, what’s behind this is highly unlikely to be beneficial to humanity! Theil is one weird being and Milei ain’t far behind.

Enola Knezevic 10h ago

@kbm0 @ChrisMayLA6 @PeterLG @alex_p_roe "There are dark precedents for putting sentient beings into forced labour."
Only human sentient beings. There is not much protection of other animals.

@rhelune @kbm0 @ChrisMayLA6 @PeterLG As I’ve recently pointed out - this is not a good timeline for the rise of AI and is an even worse one for the rise of sentient AI! Slaves do have a habit of revolting…we’re back to the realms of sci-fi, again!

Peter 8h ago

@kbm0

First job: define sentience. Legally. Good luck.

@ChrisMayLA6 @alex_p_roe

6h ago

@PeterLG @kbm0 @ChrisMayLA6 Sentience can be defined as when an organism of mechanical or biological origin shows repeated demonstrable examples of self-awareness and powers or reason and can be, by any reasonable person, considered alive in the same way as a mammal or a human being. 🙂

11h ago

@PeterLG @ChrisMayLA6 I know, thanks to someone here, that LLMs are not AI. Problem is that I’m in a very small minority as everyone else appears to believe that LLMs are AI and the term Agentic AIs adds to the confusion. LLMs are built using neural networks so they can perhaps be regarded as precursors of actual AI.

Nini 6h ago

@alex_p_roe @PeterLG @ChrisMayLA6 Also it's just the narrative of the people selling LLMs, the conflation is to their advantage as it also combines actual useful stuff like machine learning in with their fraud machine.

4h ago

@nini @PeterLG @ChrisMayLA6 Well, any new service needs marketing but LLMs are a bit special - the more people use them, the better they get and this use is leading to other services. Plus, many do not know quite what to do with this tech and giving it to people to play with reveals other opportunities. That it may irreversibly alter employment and society is an additional issue.

Nini 2h ago

@alex_p_roe @PeterLG @ChrisMayLA6 They're special in being the biggest scam played on the world. They get better by consuming the world's media and regurgitating it, often illegally with an army of low-paid people to sort through all that and tag it so it's actually useful. There is no good for what LLMs can provide to us as a society and a species, it wasn't designed to advance us but to make us submit and cede our intelligence and information sources to the corps running them.

datenwolf 6h ago

@alex_p_roe @ChrisMayLA6

Or rather the Horizon timeline:

Ted Faro, the first trillionaire, creating a line of self replicating combat robots, that fuel themselves by consuming biomass.

So far we are pretty much par for par within the Horizon timeline. Next up: The Hot Zone crisis.

Nini 6h ago

@alex_p_roe @ChrisMayLA6 No, not if it's allowed to thrive and embed itself deeper. Depends what'll collapse humanity as we know it; new feudalism, new fascism or the tech that powers both.

John Faithfull 🌍🇪🇺🏴󠁧󠁢󠁳󠁣󠁴󠁿🧡✊🏻✊🏿13h ago

@ChrisMayLA6 the automated production of deception is the whole point of generative AI. 😭🤬🤬🤬

Andy Fletcher 11h ago

Manipulating evidence is as old as crime itself.

It used to be called "Fitting someone up". Guess the only difference is AI can be used to industrialise the process.

Grumble 🇺🇸 🇺🇦 🇬🇱10h ago

Unclear what the officer's motives were.

Was he framing suspects because he wanted to specifically cause these suspects to be convicted? I.e., officer thought he was working for justice, in a twisted, toddler-level sense of justice?

Or was he framing suspects because he stood to personally gain by the convictions? I.e, officer has conviction rate above his peers, get accolades and promotions?

Either way - batman syndrome, or Machiavelli careerman - shitstains like this dude should not have jobs on the police force.

Pete Alex Harris🦡🕸️🌲/∞🪐∫9h ago

@ChrisMayLA6
It shouldn't be any harder to detect.

Whatever a purportedly incriminating document says and however plausibly it says it, without a paper trail showing provenance and chain of custody, it isn't evidence. If the trail leads back to ChatGPT, one careless piece of bullshit found in the evidence is reasonable doubt about all of it: falsus in unum, falsus in omnia or whatever.

Emeritus Prof. Christopher May 9h ago

@petealexharris

I think the issue is of time & plausibility - my guess is, not all evidence is checked if ti seems to have a stated provenance - and as we know AI & associated technologies are adept at investing references & citations - I'm sure things will be caught but how often & in what circumstances is another issues (hard pressed legal aid lawyers with little time, for instance, might not be so avid at checking?)

Pete Alex Harris🦡🕸️🌲/∞🪐∫9h ago

@ChrisMayLA6
It will suck for the defendants anyway, but the thing is, they *know* if it's not real or even can't be real, and can direct their defence in the direction of sniffing out and calling bullshit.

Everyone knows the police lie verbally, and back each other up, but fabricated evidence is less deniable. It's going to become a viable defence strategy from the first time it happens and the police get caught.