Marvellous in the proper meaning, something to marvel at:
AI promises incredible benefits, but also terrible risks. It’s not luddism to rein it in
Sonia Sodha
How?
All you need for AI is some code and some computing run time. How can you – how can anyone – limit access to those two so as to rein it in? There is absolutely no way that that is globally possible. Therefore musing over whether we should do it is a nonsense.
The optimists point out that history has proved the doomsayers wrong countless times. Take the printing press: the 15th-century Catholic church worried that the spread of information would undermine authority and stability across Europe;
Ok, how was the Pope wrong?
Butlerian Jihad!
This is one of those things that reinforced my view that journalists and fund managers really don’t know what they’re talking about. Both jobs are about confidence. Because I work in software, I’m good at math, I know what Bayesian networks and Markov chains are. I have a vague idea about how a lot of image recognition works in terms of edge detection. I’ve tinkered with NLP services like Amazon Comprehend. And I can’t say with much confidence where AI is going. If I had to guess, not actually that far.
I think in truth there’s a lot of ignorant people who have spent 5 minutes with ChatGPT, who are either doom-mongers who think anything new is a threat, or the dreamers who think it’s going to mean AI can do everything for us.
AI make me quite nostalgic for the Millenium Bug.
The arguments in that article were the same ones my father’s generation made about the spread of computers in the ‘70s.
A pencil could be used by terrorists, BAN PENCILS!
A telephone could be used by terrorists, BAN TELEPHONES!
Postal services could be used by terrorists, BAN POSTAGE!
Typewriters could be used by terrorirsts, BAN TYPEWRITERS!
Computers could be used by terrorists, BAN COMPUTERS!
She was an adviser to Milipede E. Do we need to know anything more about her? Unless she didn’t in fact advise and left the tossspot to his own devices?
The current hysteria over AI seems to be its impact on the so-called creative industries which will be unable to plagiarise, ‘reinterpret’ or apply the various tropes to reinforce lefty stereotypes
BBC output seems to have been captured about two decades ago
Steve: No over-arching authority and a modicum of instability is what we want. That’s what made The Enlightenment.
Harry Lime wasn’t wrong.
TG – but we didn’t end up with less overarching authority and more stability, did we?
We got the Prince’s Revolt, the Thirty Years War, the British Civil War, etc.
I quite like this new printing press invention, but the reactionaries were right.
People yapping on about it overlook there is a reason why it is called ‘artificial’ intelligence.
Artificial: made by Human skills; imitation; not natural; lacking in spontaneity.
The notion it can go off and do its own stuff and take over the World is the usual mix of ignorance, weakness if intellect and idiocy.
However, if it could do that, I can only seeing it as an improvement on the Human ‘intelligence’ – such as it is – running things right now.
@Steve
“ Ok, how was the Pope wrong?”
Because the Pope & Catholic Church lost its authority, power and control as the Bible and religious texts were translated from Latin into the vernacular and spread to masses via the printing press, causing widespread dissent and religious conflict.
“The current hysteria over AI seems to be its impact on the so-called creative industries which will be unable to plagiarise, ‘reinterpret’ or apply the various tropes to reinforce lefty stereotypes”
I think its more that the people who thought that they were unable to be ‘outsourced’ in the way all those blue collar workers jobs were sent overseas now realise that AI may replace their nice secure well paid jobs. And they’ll face the same dismal future they imposed on the working classes. Which of course can’t be allowed to happen…….
John B – but that’s an example of the Pope being proved correct.
So, The optimists point out that history has proved the doomsayers wrong countless times. is bollocks, intit?
The pessimists were right beyond their worst nightmares. So this isn’t a good reason to assume everything will be fine with AI.
As I understand it – correct me if I’m wrong – an AI bot is a regurgitation engine.
People who make a living from mere regurgitation might well lose income accordingly. Hurray!
“the 15th-century Catholic church worried that the spread of information would undermine authority and stability across Europe” Precision required.
The 15th-century Roman Catholic church worried that the spread of information would undermine the Pope’s authority and religious tyranny across bits of Europe.
Obviously the Orthodox Catholic church wouldn’t share those fears though it might have had fears of its own..
– The pessimists were right beyond their worst nightmares.
The sack of Rome was certainly a grim catastrophe, but to be fair there was no shortage of war and chaos before the printing press. At most it gave a new angle to justify conflict. Whether it’s the papacy, Kodak or Nokia, the more tenaciously incumbents attempt to retain their dominance in the face of a disruptive technology, the worse their beatings will be.
– As I understand it – correct me if I’m wrong – an AI bot is a regurgitation engine.
I think AI has some pattern recognition involved, and some “learning” ability to “predict” the next part of a pattern. Buggered if I know really, but it does seem a little short of prime time. If you Google (specifically Google) “African country beginning with K” you can see the results of an AI powered search engine selecting the “best” result from AI generated content. Still valid at posting of this comment despite the hilarity being up for days now.
What silly bunts.
“All you need for AI is some code and some computing run time. How can you – how can anyone – limit access to those two so as to rein it in?”
Very simply – all you need is a global police state.
This is just another person who supports such a thing because they imagine they will be the one wearing the boot.
@Steve – “how was the Pope wrong?”
The Pope was wrong in thinking that it is better to be stable and wrong than unstable and sometimes right. In particular, that the rule of the Catholic Church was based on truth and was good. Doomsayers are frequently right when you understand that their prophesy of doom is not doom for all of us – just them and their way of life.
If you think stability is such a wonderful thing, remember that you’re at your most stable once you’re dead.
“AI” is a meaningless term used by journalists and publicists to describe any computer based operatioj that they want to hype.
“AI” is a meaningless term used by journalists and publicists to describe any computer based operatio
j that they want to hype.n that they don’t understand.FTFY
Most everyone who wishes to be considered a thinker has already opined on the existential peril that is AI. However, until Paul Ehlrich pronounces I’m going to reserve judgement. You never know, he may be all in favour.
‘Doomsayers are frequently right when you understand that their prophesy of doom is not doom for all of us – just them and their way of life.’
What an accurate description of my attitude to asylum seekers, Charles.
I’m all for more and more AI.
Sooner or later someone in the New York Times or Washington Post is going to realize that if they feed their “news” output for the past fifty years into an AI system they won’t need any of those pesky “journalists” anymore.
All of the whining and caterwauling will be music to my ears.
Learn to code indeed assholes!
The idea that someone is just going to cobble together an AI in their garage computer is ridiculous.
Coding is hard.
I’m currently learning C#, and I have experience in C++ and Visual Basic. Spent hours the other day trying to figure out why my text wouldn’t change colour on the interface window, kept getting compiler errors. Turns out the auto suggested complete command of System.windows.brush.red needed to be system.windows.brushes.red.
Stack overflow is helpful, but quite a few answers are out of date.
On top of the difficulty of actually getting code to work, there’s also the fact we don’t understand how intelligence works. Sure, pattern recognition is a part of it and computers are good at that, to an extent.
Plenty of stories about how shit it actually is.
See the drones that catch a ping pong ball every time it’s thrown, zipping across a large room to get under it, then they throw a tennis ball and it just sits there.
Or the Tesla, self driving cars that look out for objects, very clever… Except when that one drove straight into a small jet at an airport and kept going because nobody had told it that planes are a thing so it didn’t have the pattern memory to look for.
Or the one that they were teaching to tell the difference between dogs and wolves. Showed loads of photos labelled as wolf or dog. When tested, they showed a picture of something that was clearly a dog, yet it always labelled as wolf. On checking to see what it was looking at to decide this, they found that in every picture they had used to teach the difference, there was snow in the background for the wolf, so instead of looking at the dog/wolf it was just looking for snow and saying wolf if there was some.
Computers are remarkably stupid. There’s no intellect, no intelligence. There is only the simulation of intelligence. As soon as they come across something they aren’t programmed for, they fall over.