https://news.sky.com/story/elon-musks-m ... s-11638381An artificial intelligence system developed by Elon Musk's OpenAI organisation is too dangerous to be released, the group believes.
OpenAI is a non-profit research organisation founded in 2015 with $1bn in backing from Mr Musk and others to promote the development of artificial intelligence technologies that benefit humanity.
The system its researchers have developed, officially called GPT-2, can generate text as it would naturally occur in language and has been released in part.
However researchers are withholding the fully-trained algorithm "due to our concerns about malicious applications of the technology".
Fake news is about to become very realistic
Moderator: Peak Moderation
Fake news is about to become very realistic
-
Little John
Load of old bollocks. It's just being hyped up, probably orchestrated by Musk.
The fact is, the system is nowhere near as clever as people are making out and, indeed, AI experts have been saying recently that they are starting to hit a wall in terms of this kind of intelligence. That is not to say, however, that there is not a lot further to go in terms of what AI is already good at becoming ever more refined.
The fact is, the system is nowhere near as clever as people are making out and, indeed, AI experts have been saying recently that they are starting to hit a wall in terms of this kind of intelligence. That is not to say, however, that there is not a lot further to go in terms of what AI is already good at becoming ever more refined.
It is already replacing lawyers and book-keepers.Software could easily replace judges,lawyers, politicians and doctors.
Sports results news articles are now written by software in many papers.
It is assisting - and will soon replace some - medical specialists.
Sentences are being determined by software and not the judges in many (US) cases.
Warehouses are now becoming the home for robots and not people.
Software and AI are ALREADY displacing jobs today ... and we are effectively at Year Zero.
Anti-technology cynicism or wishful thinking won't stop the deluge.
Standing to one side or opting out isn't an option either ... unless you fancy living in remote woodland living off the land.
-
Little John
Nobody has said it is not replacing jobs. However, the jobs you have just described are ones where there are clear rules involved. This is where AI works well. and yes, in a modern, industrial civilization, much activity is, wherever possible, automated since that is deemed as more efficient. Thus, AI can be deployed in may areas of industrial society, in turn displacing many jobs. Nobody is denying that reality.
Where AI does not work well and is showing signs of hitting hard limits is in areas where there are no clearly defined rules and where situations are highly fluid and unpredictable in real time. And nowhere is this more pronounced than in the generation of real time linguistic exchanges and output.
AI is still based, at heart, on automation based on if/then/go to algorithms no matter how abstracted those algorithms have become. And, yes, if they become sufficiently abstracted, then in terms of acting as if they are making real time "intelligent" decisions, they can look like a duck and act like a duck. But, that has its limits. Because they are still not a duck.
The one area you have cited, above, where I would be interested to see the quality of output is sports results news articles. Do you have any links?
The problem with AI fetishism is twofold:
1) Hyperbolic claims are made for it, in terms of what it is capable of or is likely to become capable of, that are entirely unproven
2) AI, being a highly complex technology inside complex industrial civilizations, sits atop an ocean of other economic activity. All of which sits atop a continual inward stream of industrial quantities of energy and other resources. Take away the stream of energy and resources and all that sits atop them will collapse in short order. That especially includes something as abstracted as AI technology.
Where AI does not work well and is showing signs of hitting hard limits is in areas where there are no clearly defined rules and where situations are highly fluid and unpredictable in real time. And nowhere is this more pronounced than in the generation of real time linguistic exchanges and output.
AI is still based, at heart, on automation based on if/then/go to algorithms no matter how abstracted those algorithms have become. And, yes, if they become sufficiently abstracted, then in terms of acting as if they are making real time "intelligent" decisions, they can look like a duck and act like a duck. But, that has its limits. Because they are still not a duck.
The one area you have cited, above, where I would be interested to see the quality of output is sports results news articles. Do you have any links?
The problem with AI fetishism is twofold:
1) Hyperbolic claims are made for it, in terms of what it is capable of or is likely to become capable of, that are entirely unproven
2) AI, being a highly complex technology inside complex industrial civilizations, sits atop an ocean of other economic activity. All of which sits atop a continual inward stream of industrial quantities of energy and other resources. Take away the stream of energy and resources and all that sits atop them will collapse in short order. That especially includes something as abstracted as AI technology.
I spent 3 months building a dedicated AI system to see if the media hype was justified or not. I can assure you that Deep Learning works. It can be applied to a huge range of tasks - enough to change our world. I suspect that maybe 80% of even mid-level jobs will be lost to AI ... or more precisely, 80% of jobs in any specific niche. This will leave the other 20% to 'manage by exception'.1) Hyperbolic claims are made for it, in terms of what it is capable of or is likely to become capable of, that are entirely unproven
AI technology hardware is simply computer memory and processors - very similar to the technology in your watch, phone or fridge. A major collapse would kill the whole lot, so you can't treat AI as a special case in this respect.2) AI, being a highly complex technology inside complex industrial civilizations, sits atop an ocean of other economic activity. All of which sits atop a continual inward stream of industrial quantities of energy and other resources. Take away the stream of energy and resources and all that sits atop them will collapse in short order. That especially includes something as abstracted as AI technology.
-
Little John
I am not treating it as any more special than a pocket watch in terms of its direct energy and resource requirements. But, by the way, such an acknowledgment, on your part, of a lack of uniqueness is not a good argument to make for its robustness as a technology in the Long Emergency to come. Rather, I am making the case for its vulnerability because of all of the interconnected, interdependent and underlying civilizational infrastructure that it relies on for its existence and ongoing maintenance which, themselves, rely on a continual inward stream of energy and other resources.
So, for example, if I grow cabbages and there is a major civilisational collapse, I can still grow cabbages because the underlying civilisational infrastructure requirements for cabbages to exist, when push comes to shove, both directly and indirectly, are relatively low. Concomitantly, in such a collapse scenario, cabbages will continue to have a direct and immediate impact on the well being of the humans who produce them and so the imperative to keep producing them and to keep in place what little infrastructure is required to maintain their production will remain high.
This analysis can be applied to literally anything that exists that is made and maintained by man.
Where do you think AI lies on the complex, civilisational hierarchy of dependencies?
So, for example, if I grow cabbages and there is a major civilisational collapse, I can still grow cabbages because the underlying civilisational infrastructure requirements for cabbages to exist, when push comes to shove, both directly and indirectly, are relatively low. Concomitantly, in such a collapse scenario, cabbages will continue to have a direct and immediate impact on the well being of the humans who produce them and so the imperative to keep producing them and to keep in place what little infrastructure is required to maintain their production will remain high.
This analysis can be applied to literally anything that exists that is made and maintained by man.
Where do you think AI lies on the complex, civilisational hierarchy of dependencies?
-
Little John
-
kenneal - lagger
- Site Admin
- Posts: 14287
- Joined: 20 Sep 2006, 02:35
- Location: Newbury, Berkshire
- Contact:
Depends which butterfly you are talking about. Cabbage white are a problem but The Red Admiral caterpillar makes an awful mess of nettles which are a problem. But then they can be a food source as well but are at their best before the caterpillars hatch so not quite so much of a problem.
Action is the antidote to despair - Joan Baez
-
Lord Beria3
- Posts: 5066
- Joined: 25 Feb 2009, 20:57
- Location: Moscow Russia
- Contact:
Ha haLittle John wrote:They'd be growing cabbages, I assure youVortex2 wrote:Just wondering : I wonder if our 'technology rich' young people would regard tech equally as important as cabbages?
Following/during a collapse perhaps huge efforts would be made by many to retain/reclaim core hi tech?
Unintentionally hilarious thread.
Geeks starving in the field. The wise one states "hey, screw growing food, let's get the computer working so that we can get ai back!"
Peace always has been and always will be an intermittent flash of light in a dark history of warfare, violence, and destruction
I suspect that many here are secretly looking forward an instant hard collapse so that they can return to the Middle Ages.Geeks starving in the field. The wise one states "hey, screw growing food, let's get the computer working so that we can get ai back!"
In reality any collapse will be much more nuanced.
As Heraclitus, the Greek philosopher said: “No man ever steps in the same river twice, for it's not the same river and he's not the same man.�
Art, literature - and yes AI & other hard won technology - will be preserved, possibly at the cost of human lives.
-
RenewableCandy
- Posts: 12780
- Joined: 12 Sep 2007, 12:13
- Location: York