Understanding the cognitive biases of AI risk skepticism
Top AI inventor Geoffrey Hinton reluctantly concluded that AI will probably humanity fairly soon
by slimboyfat 78 Replies latest social current
-
-
mikeflood
Want to check a scary movie, with all of that hype about AI?
Watch The Landmower Man from 1992....
-
slimboyfat
A good natured downplaying of AI risk by Robin Hanson here, but ultimately not terribly convincing!
https://youtu.be/aUfw7V_Wy4Y -
slimboyfat
Professor Roman Yampolskiy is the best explainer I’ve found on the topic so far. Surprisingly the death of all humans might not actually be the worst conceivable outcome from strong AI. Check out his comments around the 10 minute mark. 😮
-
Diogenesister
I thought this post, particularly the last paragraph, under a video SBF posted on this thread, very interesting! (I hope I've copied it accurately since cut and paste of replies under you tube isn't possible)
@alexshenderov4975. 6 days ago The claim that believers in Al Almighty (benevolent or otherwise) are the real experts that have thoroughly debunked every doubt expressed by those pesky skeptics makes one somewhat skeptical about the other claim: one of an open-minded discussion. Once it is assumed that flops equal intelligence, the rest of the alleged "discussion" appears rather pointless.s-wise, all the brilliant human minds at Bletchley Park - Max Newman, Andy Wiles, my Flowers, Alan Turing - couldn't hold candle to Colossus: they couldn't decrypt coded messages anywhere near as fast as Colossus did. All they did was build it and asked it the right questions. If you count s as smarts, computers were born smarter than humans. Singularity occured ;8th February 1944.rick-hauling capacity is misleading, fusing and ultimately meaningless. s comparing the human mind with a digital calculator on the basis of making digital calculations. Yet it appears to be the favourite pastime of some alleged experts.
Brick-hauling capacity is misleading, fusing and ultimately meaningless. s comparing the human mind with a digital calculator on the basis of making digital calculations. Yet it appears to be the favourite pastime of some alleged experts. Unsupported, read religious, rhetoric Al systems that are generally smarter 1 humans" is IMHO anything but benign. It is already doing serious damage to human confidence (already low in the middle of the pandemic of misanthropy & cynicism), paralyzing decision-making until a voice from the sky tells us mere mortals what to do). The prediction of human inferiority may therefore become a self-fulfilling prophecy - not by machines acquiring goal-setting capability, but by us humans forfeiting it (at a particularly bad moment, too - there are a lot of important decisions to be made ASAP). If that happens arithmetic will actually become Add a reply...
Desperation of some folks for attention other perks, but shell games with ds, substituting meanings when no s looking, don't help understanding nature of Nature). Nor are there theoretical models of how goals, bts, dreams, disputes, frustrations, osity, ambitions could emerge in a ng-computable model set run on a modular, digital, electronic hardware (again, shell games and other marketing tricks don't count). Sure, there are experts willing to stretch the interpretation of some really odd texts and drawings as far as it takes to claim the capture of a unicorn, - just like other experts stretch the interpretation of anything in Nature into evidence of an oversized white-robed bearded white male unsupported, read religious, rhetoric Al systems that are generally smarter 1 humans" is IMHO anything but ign. It is already doing serious damage uman confidence (already low in the dle of the pandemic of misanthropy cynicism), paralyzing decision-making until a voice from the sky tells us mere mortals what to do). The prediction of human inferiority may therefore become a self-fulfilling prophecy - not by machines acquiring goal-setting capability, but by us humans forfeiting it (at a particularly bad moment, too - there are a lot of important decisions to be made ASAP). If that happens arithmetic will actually become Add a reply...
Meanwhile, as transhumanists worship mythical AGI and neo-luddites feed apocalyptic scenarios to Hollywood, pragmatists are busy using machine learning. The machine does the deterministic, rational, traceable heavy lifting: crunches massive amounts of data. We the somewhat irrational humans provide imagination and direction: ask the questions, tweak the databases, evaluate the answers. This arrangement actually works in drug development and city planning, movie casting and airport security, logistics and manufacturing. Al is a useful tool, provided proper adult supervision. Hence the attempts toAl is a useful tool, provided proper adult supervision. Hence the attempts to disclaim responsibility for the future of our species, by transhumanists and neo-luddites alike, are reminiscent of earlier worldviews delegating this responsibility to Baal, Odin or Zeus. The superiority of a glorified calculator at calculations is no better reason to entrust it with making decisions for us than, say, hammer's superiority at hitting nails. It's still up to us to choose where the nails go.
Apologies this is a disaster please refer to below slims first video post!!
-
slimboyfat
I can’t follow what the person’s point is. That AI doesn’t meet his standard of intelligence? It is already passing law and medical exams and it’s only getting better from here on in. What does it have to be able to do before it’s considered a threat? The experts think it’s not far off outsmarting humans already and they are worried. Those who know most about AI are the most worried. Those who know the least are the most flippant about it. (A real life demonstration of the Dunning-Kruger effect) This wasn’t always the case because in the past some AI experts were highly skeptical about it exceeding human level intelligence any time soon. Many like Geoffrey Hinton have changed their mind due to the surprisingly rapid progress of AI in recent months. Anything by Rom Yampolskiy on the topic is worth watching.
-
slimboyfat
Just when you thought things couldn’t get any weirder, an oddly good version of “Take On Me” by “Donald Trump”. If this isn’t a sign of the end then I don’t know what is. 😂
-
Diogenesister
Brilliant version!!!! Quite prescient,no???
I beleive AI is a huge threat to humanity. I think AI will prioritise life on earth over humanity. As would any intelligent being.
-
Diogenesister
Slim Professor Roman Yampolskiy is the best explainer I’ve found on the topic so far. Surprisingly the death of all humans might not actually be the worst conceivable outcome from strong AI. Check out his comments around the 10 minute mark. 😮
Oh lor.....the JW eternal paradise!!🤪😂😂