Essay by Eric Worrall
In addition to a £31 million “Trustworthy AI” grant.
University of Southampton receives £15m AI funding from government
By Curtis Lancaster
BBC News
A university has been given £15m for a training centre aimed at developing artificial intelligence to tackle climate change.
The government funding was awarded to the University of Southampton in the hope of training at least 70 PhD students.
They will learn to use an AI technology that works on sustainability, called SustAI, the university said.
It was also awarded £31m in June towards developing trustworthy AI.
…
Read more: https://www.bbc.com/news/uk-england-hampshire-67277518
The university press release;
Southampton awarded millions to lead Britain’s AI revolution
Published: 31 October 2023
Southampton will be at the forefront of government plans to make the UK a leading force for artificial intelligence after it was awarded millions of pounds.
A new £15million training centre at the University of Southampton will be tasked with nurturing British tech talent and developing AI to tackle climate change.
The funding package, which includes £8million from UK Research and Innovation (UKRI), will forge the new AI Centre for Doctoral Training in AI for Sustainability, known as SustAI.
It will train at least 70 PhD students in sustainable AI, with plans to advance the tech for use across renewable energy and reducing carbon emissions.
SustAI director Professor Enrico Gerding, from the University of Southampton, said: “Environmental sustainability is one of the greatest challenges our world is facing – and many countries are setting ambitious targets to reduce emissions and increase renewable energy production.
“AI will be key to achieving these targets and, through SustAI, we will nurture the next generation of researchers, engineers and technologists who will be trained to create a sustainable future using AI.”
The £15million SustAI centre was announced ahead of this week’s AI safety summit, held at the famed Bletchley Park estate, which intends to fuel the UK’s ambitions to be a tech superpower.
It comes just months after Southampton was awarded £31million, also from UKRI, to launch the Responsible AI UK consortium to develop trustworthy artificial intelligence.
Associate Professor Dr Lindsay-Marie Armstrong, from the SustAI team, added: “Sustainability is at the heart of the centre, both in its research and ethos. We will equip our students with the ability to transform academic research and make a real change to businesses and society.”
More than ten other training facilities across Britain were also announced by UKRI, as part of a £117million package, all aimed at developing artificial intelligence.
UKRI chief executive Professor Dame Ottoline Leyser said the UK is in a strong position to harness the power of AI to transform many aspects of our lives for the better.
She added: “Crucial to this endeavour is nurturing the talented people and teams we need to apply AI to a broad spectrum of challenges, from healthy aging to sustainable agriculture, ensuring its responsible and trustworthy adoption.”
Applications for students to enrol onto the University of Southampton centre will open soon – ready for the start of the 2024 academic year.
Get the latest updates about SustAI by signing up to its enewsletter.
Or contact the SustAI team at eg@ecs.soton.ac.uk.
Source: https://www.southampton.ac.uk/news/2023/10/southampton-awarded-millions-to-lead-uk-ai-revolution.page
I doubt applying AI to sustainability goals will achieve much, any more than applying “Quantum Computing” to climate goals will achieve.
AI is good at optimisation problems, for example the AI in your automobile satellite navigation system can figure out an optimal route to your destination much faster than a human can.
But the fundamental problem, what to do when the sun goes down and the wind stops, or what to do about the utterly predictable collapse of renewable energy generation every winter. An optimising satnav AI cannot find a route to a destination, when there is no possible route to the desired destination.
The following is an example of an AI solving an optimisation problem – in this case finding the shortest line which joins all the dots. But the AI can only solve such an optimisation problem if a solution is possible.
Pseudo-authority by selecting the training data for the AI?
They also didn’t mention which disciplines the PhD researchers were coming from.
Environmental justice warriors will dominate.
I laugh when I hear IBM’s add for “AI you can trust”…. ie “AI that tells you what you wanted to hear”.
Yes not one word on what AI is actually supposed to be used for, but there will be 70 PhD’s to do it, whatever it is. Or maybe their job is to figure that out.
How many Ph.D.’s does it take to screw up an AI system?
“How many PhD’s does it take to screw up an AI system?”
As many as you can get funding for!
The first task for the AI should be to discover the climate change.
How much money goes to the AI, and how much to those social scientists?
its to create a new priesthood- maybe they’ll wear a special uniform- if I knew how to get AI to produce images, I’d ask it to design the new uniform or robe for this new priesthood- the Doctors of Sustainable AI
If you don’t have your own intelligence, then I guess artificial intelligence is better than nothing.
Except artificial intelligence can only mimic real intelligence. That leaves sustainability out. Forty years ago I was asked in an interview what I thought of sustainability. I said I did not know what it meant. That answer has not changed. It is a deliberately opaque concept, carefully devoid of clear meaning.
For the record I am on the I side of AI. Trying to figure out the logic of complex issues. I have yet to see an AI system that could carry on a debate, or even describe one.
Well, as the son of a cattle farmer so far as I am concerned AI is Artificial Insemination. It has worked well for many years and doesn’t need lots more PhDs, so give them something useful to do.
AI is like AGW, Crypto currency, CRT, and all the other “new and catastrophic” narratives being foisted by the MSM. Heavy on hype, light on substance, and easily manipulated for the masses to follow.
Tackle climate change or subvert it further?
“Don’t complain to me; if the computer says ya gotta eat bugs then ya gotta eat bugs, alright?”
Advocacy wins again and you lose.
It sounds to me like a demonstration of Olympic-level grantsmanship, where a really well written grant application successfully concealed the fact that the applicants hadn’t a clue how they actually were going to use the grant.
Or could it be a demonstration of how they used AI to write a successful grant application?
The scariest scenario of all would be if they actually did create an AI to tackle their idea of climate change, with all of their delusional, misanthropic Malthusian delusions.
Oops, I got distracted and put in delusions twice when I meant to put ideas as the last word.
Interesting that, if you do the repeated test for the Travelling Salesman, you get a different answer each time. Is that intentional? And is it realistic? I know nothing about AI, but shouldn’t each time produce the same answer? Otherwise, what’s the point?
It generates a random set of points each time. That’s a real AI you see running in your browser, albeit a harmless one I wrote.
It seems you have to work on your implementation. In the following example, since the crossing isn’t at a right angle the marked changes to the route give a shorter path:
There is a strong component of “more-money-than-brains” in both Climate and AI research. Sounds like a win-win for grant seekers…taxpayers not so much….
It would seem to me that if the A.I. is worth anything, it would tell us NOT to use ‘renewables’ and say USE FOSSIL FUELS, but I am sure that is not how they write the code or the grants.
Anyone ever seen that nature show about how in Africa each year, as the ponds dry up the new young frogs start eating each other to survive. It seems like we are seeing it here, as wind and solar more and more are seen NOT to serve society, there is more cannibalism to get money and energy and grants. A,I. to the rescue!
Unfortunately not. In those ponds in Aftica there is a very limited ecosystem – pondweed, tadpoles and frogs, if I remember correctly. Tadpoles eat the pondweed, frogs eat the tadpoles but not the pondweed – if there are fewer frogs one year then more tadpoles survive to be frogs and if there are fewer tadpoles then more frogs starve and die. In academia there hasn’t yet reached a saturation point where cannibalism starts – more students are graduating with nonsense degrees each year, staying in academia and the media scare stories are providing more and more grant funding. Either someone needs to start a metaphorical cull or cut off the funding and watch the carnage unfold; either way it won’t be pretty but it will be necessary – this gratuitous, unsustainable money-grubbing has to stop.
How does AI decide whether human society is worth keeping ?
Obviously, it is able to review the movie The Day the Earth Stood Still and reason from that perspective.
Frighteningly, the AI being consulted is very likely to ask: “So, have you delivered the answer that Klaatu asked for just prior to departing Earth?”
AI seems to include everything from linear regression to expert systems to impenetrable neural network voodoo that not even its practitioners pretend to understand. (As far as I can tell, the more incomprehensible it is, the more likely dimwit media types are to refer to it as an “algorithm”—the thing it least resembles!)
I would love to see a joined-up explanation of just how AI is expected to help. I can believe it will be very good at doing the “demand management” Dale Vince likes to talk about (which we will hate). But it will never in a thousand years work out how light wafts and dappled sunbeams willl power the hi-tech economy we expect.
I am not a total sceptic. There are impressive things being done (e.g. protein folding and Go) and people call them examples of AI. But just claiming one’s own particular BS is “AI” won’t make it successful or impressive. I am expecting to be chronically underwhelmed.
I still haven’t gotten a satisfactory answer for where does AI get its information. I can’t see anything other than the internet or some version of the internet. The problem is that someone is controlling the information that shows up on the internet. Or perhaps there are boundaries set for what to look for ie peer reviewed studies. That is fine but again there are people controlling what studies get peer reviewed. Nope I put no stock in this AI business.
IF it doesn’t look at everything- then it’s by definition not intelligent. It’ll have to look at what climate skeptics say. If it doesn’t indicate that it’s done that- then that’s a proof that it’s a scam.
The brainwashed will be using “trustworthy” AI to sell a pack of lies..
People have been falling for the idea computers are trustworthy and authoritative since they were invented. The whole catastrophic global warming scam is based on the assumption you can trust the computer.
And that you can trust the programmers…to be correct and to have no biases. Chances of that across the board are approaching zero.
That’s the odd part, people don’t think about the programmers. Whenever someone asks if the world would be better if it was run by an artificial intelligence, I agree completely and volunteer to be chief programmer.
A large percentage of what people believe are based on false assumptions and false narratives. That is the primary way that they are getting away with the lies. Nobody is questioning the underlying assumptions. Conclusions become invalid once the house of cards is exposed.
a truly ghastly and incestuous orgy of magical thinking being reinforced by its own self-declared authority
They’re quite ingenious when it comes to putting two words and a colossal waste of money together.
AI remains glorified pattern matching
It’s GIGO once again but, this time, with computer geeks breathlessly exclaiming “…but it can’t be rubbish, it’s AI!”
Human ingenuity confuses Google Maps.
Berlin artist uses 99 phones to trick Google into traffic jam alert
This article is more than 3 years old
Google Maps diverts road users after mistaking cartload of phones for huge traffic cluster.
https://www.theguardian.com/technology/2020/feb/03/berlin-artist-uses-99-phones-trick-google-maps-traffic-jam-alert
Quackery pays well! And why not, it’s “free” taxpayer money…
When you elect morons and scoundrels this is what you get, in any country. We have it here in the USA, too, unfortunately.
As an alumnus of Southampton Uni ( graduated :2000 aged 60 first class honours) I’m delighted they’ve got the money using ‘climate change’ as a lever. AI will quickly show its a scam.
“developing artificial intelligence to tackle climate change”
I would think that it’s required to at least offer some conceptualization as to just what AI could do for climate science. It can’t give us any data. It can’t suggest what some of the unknowns might be, or could it? It will have to enlighten itself as to what the skeptics have to say. It might be able to polish some of the models and suggest how they can be tested. It will have to consider the economics of all efforts to effect the climate. That is, it’s not about solving the supposed problem but if it concludes the problem is real, it must solve it in a responsible manner. It will have to appreciate the value systems of the proponents and opponents of the hypothetical problem. It will certainly have to consider if climate change is in some ways a positive thing.
Too much to ask of AI. Is there a “general theory of AI” comparable to Einstein’s “general theory of relativity” so we would have some idea of its potential?
A.I. answer – soak the peasants!
And they always think that someone ELSE is the peasant, when will they realize that THEY are the peasants that the elite want to eliminate.
The man behind the curtain….the Great Obomba….has already instructed Joke Biden to issue rules for AI…..to make sure there is the proper “discrimination”. If there was a real Truth Detector for AI, it would not satisfy the Left which believes anything goes for the “Cause”. If AI does not help the Cause. then AI must be made to help the Cause.
AI cannot solve the renewable intermittency problem.
AI cannot improve climate models that produce a nonexistent tropical troposphere hot spot.
AI cannot undo all the past failed climate predictions—Glacier National Park still has glaciers.
AI cannot churn out 70 new PhDs.
A new waste of climate grant money on a buzzy nonstarter.
Yeah a whole lot more computer ‘scientists’ playing computer games all day long at the taxpayers expense. Just exactly what was needed.
Time for the Threse Laws of Robotics !
😀
What if AI finds that catastrophic climate change only exists in climate models?
They’ll likely constrain it to give the “right” answer.
AI is already programmed to give only politically correct answers…the idea that it can “think” on its own is pure rubbish.
If it is truly AI it will discover the whole climate change is a hoax. But we all know AI is controlled by those who are programming it just like their models. Bias in; bias out.
My strongest recommendation: ask both Chat GPT and Bard to answer the following two questions with unambiguous “yes” or “no” replies:
Q1: Will the recent £15m AI funding grant to the University of Southampton successfully develop artificial intelligence to “tackle” climate change?
Q2: Will the June 2023 funding grant of £31m to the University of Southampton successfully develop “trustworthy” AI.
The amount of hemming-n-hawing by both AI bots as they attempt to answer both questions will as insightful as would straightforward “yes” or “no” answers (which I don’t think will be provided by either bot, despite the instruction) . . . something along the lines of “My programming does not allow me to answer such a question”.
Also, I fully expect both AI bots would give different “reasoning” as they attempt to respond to the question, raising of course this ancillary question: Will a human ever know what AI bot she/he can trust to be offering a complete and correct answer?
Somebody/some group is worried about AI being trustworthy and “tackling” climate change when it hasn’t even been demonstrated that AI has an IQ (and I do specifically mean IQ in the historic sense of that term) above room temperature?
Good grief!
A truly intelligent entity with access to the data would quickly conclude that “renewables” only serve to increase total emissions and destroy natural habitats, thus are the least viable solution to “sustainability” once either minimizing greenhouse gases or maximizing biodiversity was accepted as the goal.
If AI had any ability to unravel the logic problems involved in climate (weather predictions included) we would have seen these abilities applied already. Remember what passes for evolved AI is actually hype since the seriously difficult problems for a logic machine – languages, imagination, plot twists, random coin tossing, intuitive activity etc – are all fails and fails big time.
Full marks to Gates and company for having got politicians wrapped around their fingers to keep computers as seemingly magical machines that can do things humans cannot do. The truth is humans can do lots of stuff contemporary computers will never ever be programmed to mimic. And I would guess we are as far away from improving technology as we have ever been before with complete idiots in charge who think none of us is capable of detecting lies when they are told. That’s why they love computers which do exactly what the programmer tell them to do – lies included – and bots which apparently will get you through exams. There’s nothing new in that either.
But at least we know that those selected for this group will have the correct views on Jews and Hamas as those on the UKRI selection board have posted their hatred of Jews, support for baby-beheading Hamas terrorists and for Palestine.
More than half the grant will be used to force AI to accept that the sun, planets and gravity play little or no role in Earth’s climate change, CO2 is an evil pollutant in spite of the fat that all life on Earth depends on it, and CO2 is the primary cause of every bad weather event since the public execution of all the witches.
“A new £15million training centre at the University of Southampton will be tasked with nurturing British tech talent and developing AI to tackle climate change.”
Retrying the 2000 tech boom? Internet was not a progressive, leftist place in its first 10 years, unless you consider the type of pictures that drove modem tech development to be “progressive”. People who wonder why parents would restrict their kids cell phone ap use might not understand what it meant to be the only computer-literate human in a house at age 16 during that era.
It took the left a while to capture Internet made by introverted young men, for introverted young men. “They” won’t let that one happen again.
Ask ChatGPT: Who else has been censured by the House in the past ?
The House of Representatives has censured several members of Congress in the past. Here are some examples:
…
2. Pete Rose: In 1990, the House censured Pete Rose, a former baseball
player, for betting on baseball games while he was a player and manager.
.. is £15 Million enough ?