OK I tend to be a techno-optimist, but the more you read about the latest AI the gloomier it gets. Virtually every cognitive job will be menaced within a decade or so. WTF will that do to us?
Maybe. Then again the first ever international urban planning conference in 1898 broke up early because the attendees were so disheartened by the projections of every major city in the world being buried under 9 feet of horse manure by 1930. Then something else came along.
Perhaps the most stupid contribution to the debate I've ever seen.
You think some technological advance will come along to stop AI working?
I think the point is that society changes and adapts. The 1830s gave us the swing riots, after the thresher more or less eliminated 25% of all jobs in the country at the time. Unemployment isn't at 25% today, and if it ever does rise to those levels, we can be pretty sure it isn't due to the prevalence of the threshing machine.
I suspect AI will just prove Parkinson's Law. Work will expand to fit the spaces available.
AI is fundamentally different from 19th century technology.
What you're hoping is that the amount of work to be done will expand father than the capacity of AI to do it.
That is a huge leap of faith.
AI development could stall tomorrow. We are in uncharted territory. It has been developing with great speed but there could be unknown obstacles ahead which suddenly halt it. Eg maybe we run out of data to feed them (we are supposedly close) and then they stop getting better?
I imagine we would PROBABLY find a way around that but maybe not
No one expected it to leap ahead this quick and the corollary of that is no one knows if it will suddenly stop
Personally my estimate is it will keep developing fast for long enough to transform economies but I cannot be SURE
We've always had technological evolution that has displaced people - never mind the threshers, there's a reason why there are no more coal miners or typing pools, to give more contemporary examples.
The difference is all of these revolutions largely displaced people who were at the poorer end of the spectrum. This is the first revolution I can think of that means lawyers will have to retrain as plumbers, management consultants will have to become bottom wipers in care homes, etc.
How society will react to this is anyone's guess. However, if you subscribe to the "overproduction of elites" theory, if you think things are bad now, wait until most skill/knowledge based jobs become automated...
At that point a UBI funded by a robot/AI tax would be inevitable, there are only so many plumber and care home worker jobs to go round and university educated professionals wouldn't apply for them on the whole anyway even if made redundant
Didn't it happen on her watch anyway? Or am I missing something?
Staunton was hired to sort out Horizon and was sacked by Badenoch less than a year later. He would appear to have relatively clean hands. The key thing happening on Badenoch's watch was a high profile ITV dramatisation shining an unwanted light on Badenoch's business as usual approach to victim compensation. ie Do as little as she can get away with.
OK I tend to be a techno-optimist, but the more you read about the latest AI the gloomier it gets. Virtually every cognitive job will be menaced within a decade or so. WTF will that do to us?
Maybe. Then again the first ever international urban planning conference in 1898 broke up early because the attendees were so disheartened by the projections of every major city in the world being buried under 9 feet of horse manure by 1930. Then something else came along.
Yeah, but then you realise that SORA can create a video of a guy carrying a bowl of milk across a cathedral filled with cats to make an offering to the Cat King
Looks like it will be a lot better in 5 yrs time. Shame. It was somehow nice to know that every animated film would take years to make.
Given that “will smith eating spaghetti” (remember that?) was just 1 year ago, and given that there are three or four companies lustily pursuing this - image/video generation (it’s not just openAI) then I think animation will be revolutionised quicker than that
Eg Jeffrey katzenberg - ex chair of Disney - has just said he expects 90% of animating jobs to vanish within 3 years
@CorrectHorseBattery and many others said it was only a matter of time before Johnson was caught out. Those who had watched him in London knew that it was a matter of when, not if. So I agree with the commentary above.
I do think the odds of a Labour victory were always underpriced after GE19.
Correct Horse Battery thought NOM was a certainty in 2019 because turnout was brisk in Putney
OK I tend to be a techno-optimist, but the more you read about the latest AI the gloomier it gets. Virtually every cognitive job will be menaced within a decade or so. WTF will that do to us?
Maybe. Then again the first ever international urban planning conference in 1898 broke up early because the attendees were so disheartened by the projections of every major city in the world being buried under 9 feet of horse manure by 1930. Then something else came along.
Perhaps the most stupid contribution to the debate I've ever seen.
You think some technological advance will come along to stop AI working?
I think the point is that society changes and adapts. The 1830s gave us the swing riots, after the thresher more or less eliminated 25% of all jobs in the country at the time. Unemployment isn't at 25% today, and if it ever does rise to those levels, we can be pretty sure it isn't due to the prevalence of the threshing machine.
I suspect AI will just prove Parkinson's Law. Work will expand to fit the spaces available.
AI is fundamentally different from 19th century technology.
What you're hoping is that the amount of work to be done will expand father than the capacity of AI to do it.
That is a huge leap of faith.
AI development could stall tomorrow. We are in uncharted territory. It has been developing with great speed but there could be unknown obstacles ahead which suddenly halt it. Eg maybe we run out of data to feed them (we are supposedly close) and then they stop getting better?
I imagine we would PROBABLY find a way around that but maybe not
No one expected it to leap ahead this quick and the corollary of that is no one knows if it will suddenly stop
Personally my estimate is it will keep developing fast for long enough to transform economies but I cannot be SURE
Hence the big switch to training them on synthetic data. Get a model to generate 1000s of pages, millions even - then use that to train another (see the success in Microsofts 'Orca2' model for instance).
Much like Sora being training on Unreal Engine footage (as rumoured - openai just mention synthetic data). Coupled with techniques like those Nvidia are using with their Eureka work and their 'AI' gym. So, say you need to train a robot/AI on a task - get Sora (or similar) to generate a thousand videos of someone doing the task and let a vision model+llm train your robot.
OK I tend to be a techno-optimist, but the more you read about the latest AI the gloomier it gets. Virtually every cognitive job will be menaced within a decade or so. WTF will that do to us?
Maybe. Then again the first ever international urban planning conference in 1898 broke up early because the attendees were so disheartened by the projections of every major city in the world being buried under 9 feet of horse manure by 1930. Then something else came along.
Yeah, but then you realise that SORA can create a video of a guy carrying a bowl of milk across a cathedral filled with cats to make an offering to the Cat King
Looks like it will be a lot better in 5 yrs time. Shame. It was somehow nice to know that every animated film would take years to make.
Given that “will smith eating spaghetti” (remember that?) was just 1 year ago, and given that there are three or four companies lustily pursuing this - image/video generation (it’s not just openAI) then I think animation will be revolutionised quicker than that
Eg Jeffrey katzenberg - ex chair of Disney - has just said he expects 90% of animating jobs to vanish within 3 years
The music generation is coming along quite surprisingly quickly too. Even the free demo's you can access via the likes of Suno are quite impressive.
OK I tend to be a techno-optimist, but the more you read about the latest AI the gloomier it gets. Virtually every cognitive job will be menaced within a decade or so. WTF will that do to us?
Maybe. Then again the first ever international urban planning conference in 1898 broke up early because the attendees were so disheartened by the projections of every major city in the world being buried under 9 feet of horse manure by 1930. Then something else came along.
Yeah, but then you realise that SORA can create a video of a guy carrying a bowl of milk across a cathedral filled with cats to make an offering to the Cat King
Looks like it will be a lot better in 5 yrs time. Shame. It was somehow nice to know that every animated film would take years to make.
Given that “will smith eating spaghetti” (remember that?) was just 1 year ago, and given that there are three or four companies lustily pursuing this - image/video generation (it’s not just openAI) then I think animation will be revolutionised quicker than that
Eg Jeffrey katzenberg - ex chair of Disney - has just said he expects 90% of animating jobs to vanish within 3 years
OK I tend to be a techno-optimist, but the more you read about the latest AI the gloomier it gets. Virtually every cognitive job will be menaced within a decade or so. WTF will that do to us?
Maybe. Then again the first ever international urban planning conference in 1898 broke up early because the attendees were so disheartened by the projections of every major city in the world being buried under 9 feet of horse manure by 1930. Then something else came along.
Yeah, but then you realise that SORA can create a video of a guy carrying a bowl of milk across a cathedral filled with cats to make an offering to the Cat King
Looks like it will be a lot better in 5 yrs time. Shame. It was somehow nice to know that every animated film would take years to make.
Is AI going to feed me, clothe me or keep me warm? Is it going to install a new garage door, lay a new stair carpet, service my car? Is it going to extract my dodgy molar and replace it with an implant? Is it going to pollard my willows? And that's just this week.
Or is it merely going to address the worldwide shortage of uninteresting movies?
That's a bit churlish. It takes several years for the next eg Frozen. If that can be done in five minutes then so much the better. Unless your job is too make per frame images of Disney Princesses.
@CorrectHorseBattery and many others said it was only a matter of time before Johnson was caught out. Those who had watched him in London knew that it was a matter of when, not if. So I agree with the commentary above.
I do think the odds of a Labour victory were always underpriced after GE19.
Correct Horse Battery thought NOM was a certainty in 2019 because turnout was brisk in Putney
I thought NOM was a possibility, but a small (20-30) Conservative majority a probability. I was stunned that even against a moron like Corbyn more people didn't see through the Johnson smoke and mirrors.
Notice she doesn't actually address his allegation that he was ordered to slow down payments to Horizon victims.
Her whole argument hinges on her "I wouldn't ignore the allegations" statement.
But, er, that's plainly an assertion - not a fact.
The stuff about how "the details will emerge soon enough" is odd, too. She's speaking as if she has no power in the situation, but she's the minister in charge!
He was 75(?) and about to retire when he took on the job. I don't see any benefit to him in saying anything other than the "facts" as he understands them. And he's clearly angry about what he was asked to do.
OK I tend to be a techno-optimist, but the more you read about the latest AI the gloomier it gets. Virtually every cognitive job will be menaced within a decade or so. WTF will that do to us?
Maybe. Then again the first ever international urban planning conference in 1898 broke up early because the attendees were so disheartened by the projections of every major city in the world being buried under 9 feet of horse manure by 1930. Then something else came along.
Perhaps the most stupid contribution to the debate I've ever seen.
You think some technological advance will come along to stop AI working?
I think the point is that society changes and adapts. The 1830s gave us the swing riots, after the thresher more or less eliminated 25% of all jobs in the country at the time. Unemployment isn't at 25% today, and if it ever does rise to those levels, we can be pretty sure it isn't due to the prevalence of the threshing machine.
I suspect AI will just prove Parkinson's Law. Work will expand to fit the spaces available.
AI is fundamentally different from 19th century technology.
What you're hoping is that the amount of work to be done will expand father than the capacity of AI to do it.
That is a huge leap of faith.
AI development could stall tomorrow. We are in uncharted territory. It has been developing with great speed but there could be unknown obstacles ahead which suddenly halt it. Eg maybe we run out of data to feed them (we are supposedly close) and then they stop getting better?
I imagine we would PROBABLY find a way around that but maybe not
No one expected it to leap ahead this quick and the corollary of that is no one knows if it will suddenly stop
Personally my estimate is it will keep developing fast for long enough to transform economies but I cannot be SURE
That was exactly my son's view (who knows about this stuff). I think I mentioned it here a few months ago. The rapid development may continue or there might be a halt for awhile in the near future while obstacles are overcome.
Clips of yesterday's Palestine rally again included the chant "no ceasefire, no vote" which might be electorally important, especially if it spreads from London to Rochdale whose by-election is on the 29th.
Yes. Labour will be fine in our election (they have a buffer on the buffer) but I'm a bit worried about this for Biden.
One bit of data suggests that there is more to be said. If I understand the maths correctly, Labour need a swing of 12.7 percentage points to get an overall majority. Ignore Wellingborough - special case of banana ex MP and banana candidate. At Kingswood they got a 16 point swing. On the whole those predicting the result (including PBers) thought Labour would do better. That is not safe home territory when a GE campaign - which will be horrible and dirty - hasn't really started.
NOM remains value. And why else is Rishi 7/1 (Hills) and not 66/1 to be PM after the next election?
I don’t think it’s simply a case of Kingswood being the normal swing and Wellingborough being freakish. At the real election I reckon we’re going to get some very variable swing in different types of constituency.
Kingswood is another urban fringe seat like Uxbridge. In this parliament the defences that have fared best for the Conservatives have been Old Bexley and Crayford (10% swing), Uxbridge and South Ruislip (6.7%), and Kingswood. The worst for them have been rural Southern towns and Red Wall seats.
The GE will be fascinating. There will be surprise Labour victories in unexpected places, particularly coastal towns and the countryside, and tenacious Tory holds in rurban fringe metrolands. Labour will flatline in London and the Tories not do too badly.
Rishi is relatively more popular than Boris now in London and cities and suburban areas but less popular than Boris in coastal towns, market towns and rural areas
It’s your party that’s less popular, for having inflicted that lying idiot on us in the first place. Like you were told, at the time.
It was Boris who won the biggest Conservative landslide since Thatcher in 2019, Boris who was the first Tory leader ever to win most of the redwall seats and the largest percentage of DE voters too.
When Boris left office the Tory poll rating was also higher than it is now. You can argue about Boris not being competent enough or moral enough to be PM, however he was also the biggest Tory vote winner as leader since Thatcher
None of that contradicts what I told you, then and now.
@CorrectHorseBattery and many others said it was only a matter of time before Johnson was caught out. Those who had watched him in London knew that it was a matter of when, not if. So I agree with the commentary above.
I do think the odds of a Labour victory were always underpriced after GE19.
Correct Horse Battery thought NOM was a certainty in 2019 because turnout was brisk in Putney
He never said that was why NOM was a certainty. He believed the circumstances in 2017 would repeat, with Johnson coming unstuck and Corbyn outperforming. He was wrong and has said so. I think that deserves credit instead of piss-taking as usual from you.
But then you're never admitted to being wrong about anything and never resisted the urge to lie in every post you make, so what is going to change now?
OK I tend to be a techno-optimist, but the more you read about the latest AI the gloomier it gets. Virtually every cognitive job will be menaced within a decade or so. WTF will that do to us?
Maybe. Then again the first ever international urban planning conference in 1898 broke up early because the attendees were so disheartened by the projections of every major city in the world being buried under 9 feet of horse manure by 1930. Then something else came along.
Yeah, but then you realise that SORA can create a video of a guy carrying a bowl of milk across a cathedral filled with cats to make an offering to the Cat King
Looks like it will be a lot better in 5 yrs time. Shame. It was somehow nice to know that every animated film would take years to make.
Given that “will smith eating spaghetti” (remember that?) was just 1 year ago, and given that there are three or four companies lustily pursuing this - image/video generation (it’s not just openAI) then I think animation will be revolutionised quicker than that
Eg Jeffrey katzenberg - ex chair of Disney - has just said he expects 90% of animating jobs to vanish within 3 years
The same is going to be true of special effects.
Also consider things like current A/B testing of film endings. Costly, cumbersome. Especially if you need to reshoot.
Didn't it happen on her watch anyway? Or am I missing something?
Staunton was hired to sort out Horizon and was sacked by Badenoch less than a year later. He would appear to have relatively clean hands. The key thing happening on Badenoch's watch was a high profile ITV dramatisation shining an unwanted light on Badenoch's business as usual approach to victim compensation. ie Do as little as she can get away with.
If it’s truly a disgraceful misrepresentation of the conversation then tough. Ha ha. Thump.
Considering the way Badenoch and chums in the right wing press misrepresent the views of normal people it seems fair enough.
OK I tend to be a techno-optimist, but the more you read about the latest AI the gloomier it gets. Virtually every cognitive job will be menaced within a decade or so. WTF will that do to us?
Maybe. Then again the first ever international urban planning conference in 1898 broke up early because the attendees were so disheartened by the projections of every major city in the world being buried under 9 feet of horse manure by 1930. Then something else came along.
Yeah, but then you realise that SORA can create a video of a guy carrying a bowl of milk across a cathedral filled with cats to make an offering to the Cat King
Looks like it will be a lot better in 5 yrs time. Shame. It was somehow nice to know that every animated film would take years to make.
Is AI going to feed me, clothe me or keep me warm? Is it going to install a new garage door, lay a new stair carpet, service my car? Is it going to extract my dodgy molar and replace it with an implant? Is it going to pollard my willows? And that's just this week.
Or is it merely going to address the worldwide shortage of uninteresting movies?
That's a bit churlish. It takes several years for the next eg Frozen. If that can be done in five minutes then so much the better. Unless your job is too make per frame images of Disney Princesses.
Also, yes, I expect Ai robots to be able to do many of those jobs within a decade
Install a garage door yes Lay a carpet yes Dentistry absolutely Pollarding yes
Here is a robot doing the dishes. The video is accelerated so a little deceptive but the progress is still astounding
OK I tend to be a techno-optimist, but the more you read about the latest AI the gloomier it gets. Virtually every cognitive job will be menaced within a decade or so. WTF will that do to us?
Maybe. Then again the first ever international urban planning conference in 1898 broke up early because the attendees were so disheartened by the projections of every major city in the world being buried under 9 feet of horse manure by 1930. Then something else came along.
Perhaps the most stupid contribution to the debate I've ever seen.
You think some technological advance will come along to stop AI working?
I think the point is that society changes and adapts. The 1830s gave us the swing riots, after the thresher more or less eliminated 25% of all jobs in the country at the time. Unemployment isn't at 25% today, and if it ever does rise to those levels, we can be pretty sure it isn't due to the prevalence of the threshing machine.
I suspect AI will just prove Parkinson's Law. Work will expand to fit the spaces available.
AI is fundamentally different from 19th century technology.
What you're hoping is that the amount of work to be done will expand father than the capacity of AI to do it.
That is a huge leap of faith.
AI development could stall tomorrow. We are in uncharted territory. It has been developing with great speed but there could be unknown obstacles ahead which suddenly halt it. Eg maybe we run out of data to feed them (we are supposedly close) and then they stop getting better?
I imagine we would PROBABLY find a way around that but maybe not
No one expected it to leap ahead this quick and the corollary of that is no one knows if it will suddenly stop
Personally my estimate is it will keep developing fast for long enough to transform economies but I cannot be SURE
Hence the big switch to training them on synthetic data. Get a model to generate 1000s of pages, millions even - then use that to train another (see the success in Microsofts 'Orca2' model for instance).
Much like Sora being training on Unreal Engine footage (as rumoured - openai just mention synthetic data). Coupled with techniques like those Nvidia are using with their Eureka work and their 'AI' gym. So, say you need to train a robot/AI on a task - get Sora (or similar) to generate a thousand videos of someone doing the task and let a vision model+llm train your robot.
OK I tend to be a techno-optimist, but the more you read about the latest AI the gloomier it gets. Virtually every cognitive job will be menaced within a decade or so. WTF will that do to us?
Maybe. Then again the first ever international urban planning conference in 1898 broke up early because the attendees were so disheartened by the projections of every major city in the world being buried under 9 feet of horse manure by 1930. Then something else came along.
Perhaps the most stupid contribution to the debate I've ever seen.
You think some technological advance will come along to stop AI working?
I think the point is that society changes and adapts. The 1830s gave us the swing riots, after the thresher more or less eliminated 25% of all jobs in the country at the time. Unemployment isn't at 25% today, and if it ever does rise to those levels, we can be pretty sure it isn't due to the prevalence of the threshing machine.
I suspect AI will just prove Parkinson's Law. Work will expand to fit the spaces available.
AI is fundamentally different from 19th century technology.
What you're hoping is that the amount of work to be done will expand father than the capacity of AI to do it.
That is a huge leap of faith.
AI development could stall tomorrow. We are in uncharted territory. It has been developing with great speed but there could be unknown obstacles ahead which suddenly halt it. Eg maybe we run out of data to feed them (we are supposedly close) and then they stop getting better?
I imagine we would PROBABLY find a way around that but maybe not
No one expected it to leap ahead this quick and the corollary of that is no one knows if it will suddenly stop
Personally my estimate is it will keep developing fast for long enough to transform economies but I cannot be SURE
We've always had technological evolution that has displaced people - never mind the threshers, there's a reason why there are no more coal miners or typing pools, to give more contemporary examples.
The difference is all of these revolutions largely displaced people who were at the poorer end of the spectrum. This is the first revolution I can think of that means lawyers will have to retrain as plumbers, management consultants will have to become bottom wipers in care homes, etc.
How society will react to this is anyone's guess. However, if you subscribe to the "overproduction of elites" theory, if you think things are bad now, wait until most skill/knowledge based jobs become automated...
I think you'll find that "Lawyers need to figure out ways of making it illegal to replace them".
Eg Jeffrey katzenberg - ex chair of Disney - has just said he expects 90% of animating jobs to vanish within 3 years
One of the more entertaining parts of Person of Interest is the filler sections which are supposed to show the machine's view of the world. Every moving object in the scene has a graphic around it to signify being tracked in real time by AI. I wonder how many hours in a VFX studio those clips took to create by hand...
Tories had fun trying to pin the thing on Davey. Its all his fault. What was Badenoch doing? Instructing the PO not to pay out compensation to limit the government's liability.
No point being appalled any more, is there. They can't stoop any lower.
So, you are a cabinet minister. Money is very tight and you are trying to identify savings. Someone comes to you and says that there is a group of people who have been poisoned by blood products or wrongly accused of theft as a result of utter incompetence (and a naff computer system) or have been sexually abused whilst under the care of the State.
its going to cost big money to fix it or acknowledge the problem. Your civil servants are very unhappy because they see the implications for their projects. it is likely some of the blame will attach to you, even if you try to fix it. Your estimated time in your current post is 18 months.
What do you do?
(a) state that this is a complete disgrace, address it and slash spending elsewhere. (b) Try and bury it so that the money never has to be spent. (c) Knock it down the road (inquiries suggest you are actually doing something and buy a lot of time) and leave it for the next guy to sort out.
You forgot pay your chums obscene amounts for years to chair/participate in the whitewash.
OK I tend to be a techno-optimist, but the more you read about the latest AI the gloomier it gets. Virtually every cognitive job will be menaced within a decade or so. WTF will that do to us?
Maybe. Then again the first ever international urban planning conference in 1898 broke up early because the attendees were so disheartened by the projections of every major city in the world being buried under 9 feet of horse manure by 1930. Then something else came along.
Yeah, but then you realise that SORA can create a video of a guy carrying a bowl of milk across a cathedral filled with cats to make an offering to the Cat King
Looks like it will be a lot better in 5 yrs time. Shame. It was somehow nice to know that every animated film would take years to make.
Given that “will smith eating spaghetti” (remember that?) was just 1 year ago, and given that there are three or four companies lustily pursuing this - image/video generation (it’s not just openAI) then I think animation will be revolutionised quicker than that
Eg Jeffrey katzenberg - ex chair of Disney - has just said he expects 90% of animating jobs to vanish within 3 years
The same is going to be true of special effects.
Also consider things like current A/B testing of film endings. Costly, cumbersome. Especially if you need to reshoot.
:: click :: whirr :: beep ::
"Ok, how about this version?"
You could do 100 versions of every movie and focus group them again and again and again and change them - at no cost - again and again and again - until you get the perfect audience reaction
We can expect AI to create amazing tv and movies. Just won’t involve many people in the creation
You could do 100 versions of every movie and focus group them again and again and again and change them - at no cost - again and again and again - until you get the perfect audience reaction
No
Disney made the 3rd new Star Wars movie the fans said they wanted.
You could do 100 versions of every movie and focus group them again and again and again and change them - at no cost - again and again and again - until you get the perfect audience reaction
No
Disney made the 3rd new Star Wars movie the fans said they wanted.
And it's not very good...
That's more because it was what an obsessive group of antisocial fans wanted. Rather than what may have tested well with wider audiences.
If AIs do ever achieve purposeful intelligence, they'll be able to outthink us a thousand times a second. So that putative battle is already lost.
Not if we get to the plug first.
What plug ?
Our society can't function without tech, and these developments are so widely distributed - and rapidly becoming far more so - that there isn't a plug. Or even a few thousand plugs.
OK I tend to be a techno-optimist, but the more you read about the latest AI the gloomier it gets. Virtually every cognitive job will be menaced within a decade or so. WTF will that do to us?
Maybe. Then again the first ever international urban planning conference in 1898 broke up early because the attendees were so disheartened by the projections of every major city in the world being buried under 9 feet of horse manure by 1930. Then something else came along.
Yeah, but then you realise that SORA can create a video of a guy carrying a bowl of milk across a cathedral filled with cats to make an offering to the Cat King
Looks like it will be a lot better in 5 yrs time. Shame. It was somehow nice to know that every animated film would take years to make.
Given that “will smith eating spaghetti” (remember that?) was just 1 year ago, and given that there are three or four companies lustily pursuing this - image/video generation (it’s not just openAI) then I think animation will be revolutionised quicker than that
Eg Jeffrey katzenberg - ex chair of Disney - has just said he expects 90% of animating jobs to vanish within 3 years
The same is going to be true of special effects.
Also consider things like current A/B testing of film endings. Costly, cumbersome. Especially if you need to reshoot.
:: click :: whirr :: beep ::
"Ok, how about this version?"
You could do 100 versions of every movie and focus group them again and again and again and change them - at no cost - again and again and again - until you get the perfect audience reaction
We can expect AI to create amazing tv and movies. Just won’t involve many people in the creation
A film made by machine and optimised to deliver exactly what audiences want and expect is the opposite of art.
OK I tend to be a techno-optimist, but the more you read about the latest AI the gloomier it gets. Virtually every cognitive job will be menaced within a decade or so. WTF will that do to us?
Maybe. Then again the first ever international urban planning conference in 1898 broke up early because the attendees were so disheartened by the projections of every major city in the world being buried under 9 feet of horse manure by 1930. Then something else came along.
Yeah, but then you realise that SORA can create a video of a guy carrying a bowl of milk across a cathedral filled with cats to make an offering to the Cat King
Looks like it will be a lot better in 5 yrs time. Shame. It was somehow nice to know that every animated film would take years to make.
Given that “will smith eating spaghetti” (remember that?) was just 1 year ago, and given that there are three or four companies lustily pursuing this - image/video generation (it’s not just openAI) then I think animation will be revolutionised quicker than that
Eg Jeffrey katzenberg - ex chair of Disney - has just said he expects 90% of animating jobs to vanish within 3 years
The same is going to be true of special effects.
Also consider things like current A/B testing of film endings. Costly, cumbersome. Especially if you need to reshoot.
:: click :: whirr :: beep ::
"Ok, how about this version?"
You could do 100 versions of every movie and focus group them again and again and again and change them - at no cost - again and again and again - until you get the perfect audience reaction
We can expect AI to create amazing tv and movies. Just won’t involve many people in the creation
A film made by machine and optimised to deliver exactly what audiences want and expect is the opposite of art.
OK I tend to be a techno-optimist, but the more you read about the latest AI the gloomier it gets. Virtually every cognitive job will be menaced within a decade or so. WTF will that do to us?
Maybe. Then again the first ever international urban planning conference in 1898 broke up early because the attendees were so disheartened by the projections of every major city in the world being buried under 9 feet of horse manure by 1930. Then something else came along.
Yeah, but then you realise that SORA can create a video of a guy carrying a bowl of milk across a cathedral filled with cats to make an offering to the Cat King
Looks like it will be a lot better in 5 yrs time. Shame. It was somehow nice to know that every animated film would take years to make.
Given that “will smith eating spaghetti” (remember that?) was just 1 year ago, and given that there are three or four companies lustily pursuing this - image/video generation (it’s not just openAI) then I think animation will be revolutionised quicker than that
Eg Jeffrey katzenberg - ex chair of Disney - has just said he expects 90% of animating jobs to vanish within 3 years
The same is going to be true of special effects.
Also consider things like current A/B testing of film endings. Costly, cumbersome. Especially if you need to reshoot.
:: click :: whirr :: beep ::
"Ok, how about this version?"
You could do 100 versions of every movie and focus group them again and again and again and change them - at no cost - again and again and again - until you get the perfect audience reaction
We can expect AI to create amazing tv and movies. Just won’t involve many people in the creation
A film made by machine and optimised to deliver exactly what audiences want and expect is the opposite of art.
Perhaps. But I fear it will be highly effective
Yes and no. I think it's still the writing that ultimately determines the success of films, unless the film is part of a franchise and derives its success ftom preexisting IP. Right now AI can reduce the cost of many processes (just as technology has since movies first started) but it can't come up with original ideas or really engage with the human condition. That might come but I doubt it. I mean, we are smarter than dogs but can we really make great movies for dogs?
OK I tend to be a techno-optimist, but the more you read about the latest AI the gloomier it gets. Virtually every cognitive job will be menaced within a decade or so. WTF will that do to us?
Maybe. Then again the first ever international urban planning conference in 1898 broke up early because the attendees were so disheartened by the projections of every major city in the world being buried under 9 feet of horse manure by 1930. Then something else came along.
Yeah, but then you realise that SORA can create a video of a guy carrying a bowl of milk across a cathedral filled with cats to make an offering to the Cat King
Looks like it will be a lot better in 5 yrs time. Shame. It was somehow nice to know that every animated film would take years to make.
Is AI going to feed me, clothe me or keep me warm? Is it going to install a new garage door, lay a new stair carpet, service my car? Is it going to extract my dodgy molar and replace it with an implant? Is it going to pollard my willows? And that's just this week.
Or is it merely going to address the worldwide shortage of uninteresting movies?
That's a bit churlish. It takes several years for the next eg Frozen. If that can be done in five minutes then so much the better. Unless your job is too make per frame images of Disney Princesses.
Also, yes, I expect Ai robots to be able to do many of those jobs within a decade
Install a garage door yes Lay a carpet yes Dentistry absolutely Pollarding yes
Here is a robot doing the dishes. The video is accelerated so a little deceptive but the progress is still astounding
Eg Jeffrey katzenberg - ex chair of Disney - has just said he expects 90% of animating jobs to vanish within 3 years
One of the more entertaining parts of Person of Interest is the filler sections which are supposed to show the machine's view of the world. Every moving object in the scene has a graphic around it to signify being tracked in real time by AI. I wonder how many hours in a VFX studio those clips took to create by hand...
Very few. Using a computer to trace a spline between two keyframes and moving them smoothly between frames has been around since the mid-90s.
Comments
Eg Jeffrey katzenberg - ex chair of Disney - has just said he expects 90% of animating jobs to vanish within 3 years
Much like Sora being training on Unreal Engine footage (as rumoured - openai just mention synthetic data). Coupled with techniques like those Nvidia are using with their Eureka work and their 'AI' gym. So, say you need to train a robot/AI on a task - get Sora (or similar) to generate a thousand videos of someone doing the task and let a vision model+llm train your robot.
https://blogs.nvidia.com/blog/eureka-robotics-research/
https://www.youtube.com/watch?v=lyaoewyXAxM
https://deepmind.google/discover/blog/scaling-up-learning-across-many-different-robot-types/
https://www.suno.ai/
But then you're never admitted to being wrong about anything and never resisted the urge to lie in every post you make, so what is going to change now?
Leave Horse alone.
:: click :: whirr :: beep ::
"Ok, how about this version?"
If it’s truly a disgraceful misrepresentation of the conversation then tough. Ha ha. Thump.
Considering the way Badenoch and chums in the right wing press misrepresent the views of normal people it seems fair enough.
Install a garage door yes
Lay a carpet yes
Dentistry absolutely
Pollarding yes
Here is a robot doing the dishes. The video is accelerated so a little deceptive but the progress is still astounding
https://x.com/chichengcc/status/1758539728444629158?s=46&t=bulOICNH15U6kB0MwE6Lfw
We can expect AI to create amazing tv and movies. Just won’t involve many people in the creation
Disney made the 3rd new Star Wars movie the fans said they wanted.
And it's not very good...
https://en.wikipedia.org/wiki/Inbetweening