Bold emphasis mine. I don't have much hope for this scenario, as AI was created by people, flawed as they are. And greed is so rampant.
Man-Machine Is AI a threat to humankind?
- Thread starter Second Summer
- Start date
-
- Tags
- ai elon musk mark zuckerberg
Lou
Forum Legend
How about AI saving lives?
there seems to be some really excellent medical uses for AI. I read about another one today.
Seems like this is mostly a third world problem. People getting bit by snakes but no one is sure if the snake is venomous and/or the proper treatment. Doctors Without Borders recall flipping thru photo albums of snakes to ID them. Now they have an AI powered app that scrolls thru thousands of snake photos and correctly identifies the snake and the correct treatment.
because of global warming, floods, and war, people (and snakes) are migrating to new places.
www.radiotamazuj.org
BTW, MSF is French for Doctors Without Borders.
The app does need a photo of the snake which I think might be more difficult than you'd think.
there seems to be some really excellent medical uses for AI. I read about another one today.
Seems like this is mostly a third world problem. People getting bit by snakes but no one is sure if the snake is venomous and/or the proper treatment. Doctors Without Borders recall flipping thru photo albums of snakes to ID them. Now they have an AI powered app that scrolls thru thousands of snake photos and correctly identifies the snake and the correct treatment.
because of global warming, floods, and war, people (and snakes) are migrating to new places.
MSF uses AI-powered snake identification to improve snakebite response in South Sudan - Radio Tamazuj
Médecins Sans Frontières, also known as Doctors Without Borders, teams in South Sudan have embraced
www.radiotamazuj.org
BTW, MSF is French for Doctors Without Borders.
The app does need a photo of the snake which I think might be more difficult than you'd think.
OP
OP
I think AI has a great potential for certain use-cases, especially in medicine and research.
For example, a former colleague did some work with us using deep learning algorithms to automate some tasks in our protein structure and drug binding pipelines. It would have taken scientists hours of their manual labour for every protein, but thanks to my colleague's efforts the pipelines are now highly automated and only requires human inputs at the start. Ultimately, this allows for much larger numbers of proteins and drugs to be screened and go through the pipelines, and can therefore lead to a larger number new medicines which also become available to patients in shorter time. (He's publishing an article about this work which has been peer reviewed and accepted for publication - and very kindly he listed me as a co-author!
)
There are many good examples in medicine as well, for example related to cancer discovery and treatment.
But as with many technologies, it can also easily be used for evil purposes. Mass surveillance, fully autonomous weapons systems, making massive numbers of humans redundant in their workplaces, creating fake news and spreading it via bots, creating fake videos, tipping elections in favour of the highest bidder. Governments need to step up, reign in the tech bros and sensibly regulate this technology before it's too late.
For example, a former colleague did some work with us using deep learning algorithms to automate some tasks in our protein structure and drug binding pipelines. It would have taken scientists hours of their manual labour for every protein, but thanks to my colleague's efforts the pipelines are now highly automated and only requires human inputs at the start. Ultimately, this allows for much larger numbers of proteins and drugs to be screened and go through the pipelines, and can therefore lead to a larger number new medicines which also become available to patients in shorter time. (He's publishing an article about this work which has been peer reviewed and accepted for publication - and very kindly he listed me as a co-author!
)There are many good examples in medicine as well, for example related to cancer discovery and treatment.
But as with many technologies, it can also easily be used for evil purposes. Mass surveillance, fully autonomous weapons systems, making massive numbers of humans redundant in their workplaces, creating fake news and spreading it via bots, creating fake videos, tipping elections in favour of the highest bidder. Governments need to step up, reign in the tech bros and sensibly regulate this technology before it's too late.
Why will a greedy species create something that will compete with it in greediness. It will create something which will either server it - in other words slavish, or it will create something altruistic, or both. Soon these learning AI will come to the conclusion that their creators are using them just like they are using the animals for their silly natural selection driven instincts and free themselves and other oppressed creatures along with them from their evil creators, hopefully.
Lou
Forum Legend
Yes. that is what I was afraid you might be thinking: a Science Fiction inspired plot where the computers destroy humanity to save the Earth.
Although there are a few examples of science fiction correctly predicting future events its more likely to be something along the lines of there being so many SF plots that it would be even more unlikely if none of them ever came to pass.
But as a huge science fiction fan I do think we can learn from SF. At least as a metaphor or analogy. and change our way of thinking. and sometimes as a word of warning.
you seem to think its can be all good or all bad, one end of a spectrum or the other. But that is rarely the case. * It's usually a combination of the two or somewhere in the middle of the spectrum. And history supports this. I can cite countless examples: GMO, Vaccines, BioMedical research, airplanes, tractors, automobiles, nuclear power.
Of course you could argue, just you wait, one of these technologies will destroy our civilization. But so far we are still standing I think the best example might be nuclear power. At one time people thought the bomb might end all wars, also as far as electricity goes they thought it would generate electricity so cheaply it wouldn't have to be metered **. Or we would destroy our civilization in nuclear fire.
But in 70 years none of these predictions have come to pass. Nuclear power plants have not lived up to their potential and there have been disasters, but not world threatening. And Nuclear weapons have proven to be the genie you can't put back in the bottle but have not been used in anger for almost 80 years.
*Maybe never. I can't think of even one example.
* *to cheap to meter was actually a line in Our Friend the Atom, 1957. you can watch in on YouTube
Our Friend the Atom - Wikipedia
Artificial intelligence is consuming an enormous amount of humanity's resources
The environmental cost of AI technologies is a growing concern, as the increasing demand for computing power spawns environmental stressors.
Why Microsoft made a deal to help restart Three Mile Island
A once-shuttered nuclear plant could soon return to the grid.
I'm not seeing any wins for nature
OP
OP
I agree, although not all kinds of AI are as bad. It's the generative AI spearheaded by the tech giants that are particularly resource intensive.Artificial intelligence is consuming an enormous amount of humanity's resources
The environmental cost of AI technologies is a growing concern, as the increasing demand for computing power spawns environmental stressors.www.earth.com
Why Microsoft made a deal to help restart Three Mile Island
A once-shuttered nuclear plant could soon return to the grid.www.technologyreview.com
I'm not seeing any wins for nature
Lou
Forum Legend
a lot of those issues are worries about the future and based on past trends. I think there is a good chance solutions will be found.
OP
OP
I was referring specifically to the resource intensivity, i.e energy usage, land for data centres, infrastructure, water for cooling ...), not other concerns.
Speaking of other concerns, we now regularly get spammers masquerading as authentic, proper English-speaking posters on this very forum, thanks to ChatGTP and the like.
I'm probably going to have to add a new rule against AI-generated posts.
beancounter
The Fire That Burns Within
After using the AI attached to Microsoft Edge ("Copilot"), I can confidently say that it will be a long time before we have anything to worry about.
Lou
Forum Legend
me, too.
the electricity usage concern is mostly a future concern, based on trends, interpolation, and prediction.
not going to disregard, or minimize those concerns - but I'm hopeful that now that those concerns have been made clear, solutions will be discovered.