Daily podcast: How risky is artificial intelligence?
- November 26, 2023
- Posted by: MainInstructor
- Category: Artificial Intelligence Go
Video Title: Daily podcast: How risky is artificial intelligence?
We hear a lot about AI at the moment don’t we how it’s going to replace our jobs the threat it poses to humanity how it’s developing at a rate that’s alarming even those developing it but are the risks really that great and do they outweigh all the benefits I mean
It’s brought the Beatles back with a brand new song it’s a new Beatle single written by John Lennon played by John Paul George and Ringo how great is that a new Beatles release half a century after they broke up all thanks to the magic of AI but this week
World leaders and Tech experts have been trying to address all the worrying aspects of AI at the world’s first safety Summit AT Bletchley Park home to World War II’s code Breakers who cracked the Nazis Enigma machine so how’s it gone let’s find out I’m Sally Lockwood and this is the Sky News Daily first we’re going to speak to Stuart Russell an AI Guru or officially professor of computer science at the University of California Berkeley and a member of the world economic forum’s Global AI Council so AI artificial intelligence is I think most people understand it’s about making machines intelligent and
The question is what does that mean for most of the history of the field it’s meant something pretty simple that the machine’s actions can be expected to achieve its objectives so if you think about using your GPS and the objective is the destination that you want to get
To the algorithm figures out how to get there and then chooses where to turn left and turn right and all that stuff people have argued forever about whether it’s actually thinking whether it’s really having subjective experience whether it’s conscious but from a practical point of view uh we only
Really care about what it does and what it’s going to do uh and for that we don’t have to answer the question of whether it’s actually conscious and just how big a moment are we in when it comes to AI revolutionizing how we live so I
Think if you’d ask me that question a year ago I would have pointed to some important progress that had been made in recognizing objects in images and transcribing speech and other useful tasks that machines can do um and we would just have seen some of the first generative AI systems that could produce
Images so you would type in a description you know I want a picture of members of the House of Lords mud wrestling and uh and then it would produce a picture do its best to come up with an image that corresponds to that description didn’t work all that well
But it’s getting better and then chat GPT happened in November of last year and that completely changed the discourse very soon politicians and CEO as well as normal people started using chat GPT and were really taken aback by how intelligent it seemed how it was able to you know write a whole editorial
Article about the problem of Peace in Ukraine or whatever it might be and I no longer needed to explain to politicians uh why AI might represent a long-term risk to humanity uh as AI systems become more intelligent than human beings because they could feel it it was sort
Of a foretaste of a future where there’s unlimited Intelligence on tap and we’re not there yet but that foraste was enough to Galvanize the world in my view and I think most of my colleagues would say it does not yet reach the level that we would call what some people call AGI
Artificial general intelligence meaning AI systems that are as capable or more capable than humans along every relevant dimension and a group at Microsoft of very distinguished AI researchers studied it for several months a week after the release they published a paper saying that it exhibited Sparks of artificial general intelligence and I
Think that followed by the letter suggesting that we should actually slow down and figure out how to design these things properly how to regulate them really got the attention of governments now a lot of the debate around chat GPT and AI is about the existential risk it
Poses how valid is that or is that slightly exaggerated ascribing existential risk to gbt 4 I think very few people are actually doing that um they’re pointing to a number of other risks that it poses for example the ability to generate very convincing disinformation the the ability to give
Stepbystep handholding advice to a terrorist on exactly how to build a biological weapon for example example if there were to be a major catastrophe in the next 12 to 18 months as a result of AI it would probably be one or both of those things the systems that we have
Now exhibit uh significant weaknesses probably the biggest weakness is the inability to construct uh and execute long-term plans so that will be the kind of thing which would enable a system to present a real risk to humanity so let me give you an analogy right if you if
You think about playing chess against a computer um anyone who plays chess knows that computers are far far better than human beings because they can just look much further ahead into the future of the game they’re often thinking 35 or 40 moves ahead which is more than even the
Best human players could do and that means that if you try to win uh you will fail they will win and that’s the end of it uh so if you take that idea and extend it into the real world where the AI system can do that kind of
Look ahead that construct those plans that are just thinking much further ahead than humans and if they’re pursuing an objective that’s in conflict with what humans want the future to be like then who’s going to win the AI system assuming that it’s sufficiently capable so it seems inevitable that if
We don’t figure out how to make these powerful systems that will be developed and many people think they’ll be developed in the next 5 years if we don’t figure out how to make those systems guaranteed safe so we can be absolutely sure that what they do will
Not be in conflict with what humans want so that’s the problem we Face how can we ensure human control forever over machines that will become more powerful than human beings just in terms of AI developing in tandem with the right safeguards being put in place are we
There or is it too late for that I think it’s definitely not too late it’s it’s certainly true that progress is very rapid and the Investments going specifically into building artificial general intelligence are actually far greater than for example the entire science budget of the United States governments need to think about
Something Beyond voluntary uh self-regulation you know there’s a lot of history showing that you need an external uh constraint on what corporations are doing because they’re quite legitimately pursuing their interests but those interests may not be aligned with the interests of the human race and so governments are authorized
And should intervene where it’s appropriate and the kinds of regulations I think that we need right now one approach that seems to be gaining favor is that just as we do with medicines and we do in fact with nuclear power stations the developer has to present to the regulator essentially uh a proof
That the system the nuclear power station the medicine or in this case the AI system meets the required safety properties we are not anywhere close to being able to do that with AI systems so if I just asked you know the biggest companies in the world can you guarantee
That your AI system is not going to to replicate itself onto other computers without permission can you guarantee that it’s not going to break into someone else’s computer system and steal information can you guarantee that it’s not going to advise terrorists on how to build biological weapons at the moment
They would probably say no and the answer that the regulator should give that government should give is well tough come back when you can and then we’ll talk about allowing your systems to be deployed they we shouldn’t take this excuse that oh it’s really difficult normal person would say well
That’s pathetic you know you’ve got hundreds of billions of pounds and you can’t even guarantee something as simple as that so go back to the lab and get on with it and figure it out how to do it and just lastly Stuart do you feel confident after this week the direction
That AI is now moving in I am more optimistic as a result of the summit I think that the direction of the technology is still going in the wrong the wrong direction we are still very much thinking about building bigger and bigger black boxes that we don’t understand training them on more and
More data and sort of hoping for the best and in fact the way we’re training these systems mostly on data generated by humans in the form of text so uh reading newspapers scientific journal articles encyclopedias uh eventually you know pretty much anything the human race has
Ever written is going to be ingested and so what we’re training them to do is to imitate human linguistic behavior and human linguistic behavior is generated by humans and the humans had purposes in generating that behavior for example I want you to buy this car I want you to
Vote for me I want to be rich I want you to fall in love with me and marry me so what we’re actually training these systems to do is first of all to to have a multiple personality disorder because essentially the system is learning various different types of human
Language behavior and types of goal and types of personality and you can evoke different personalities by talking to the system in different ways for example I I often if my students can’t get the system to produce a useful answer to the question I said well did you say please
And you might think what do that got to do with it and in fact if you say please if you cou your question politely you’re evoking the personality that’s sort of like the helpful librarian as opposed to the personality that’s like you know the abrasive politician or the drunk in the
Power whatever it might be and it seems to work unfortunately we actually don’t want machines to have human goals right and I usually use this example of drinking coffee if I’m drinking coffee and the and the Machine is observing me drinking coffee I don’t want it to want
To drink coffee I want it to know that I like coffee and to make me a cup of coffee but I don’t want it to drink coffee so training something to copy and adopt human goals is a fundamental mistake but that’s what we’re doing with this technology once we demand that the
Developers prove that their systems are safe they’re going to have to go back to the drawing board at some extent and rethink the technological approach that they’re taking that has been an absolutely fascinating Insight Stuart I really appreciate your time thanks so much for joining us today a pleasure
After the break I’ll be chatting to our technology correspondent artti nappen who’s been taking the temperature at Bletchley Park all Week our technology correspondent artti nappen joins us now from the hustle and bustle of the global Summit AT Bletchley Park hi artti now we hear a lot about how the UK wants to become a world leader in AI hosting this Summit a big part of that is it living up to this
Reputation they’ve been making a big deal in government about the fact that this is the first ever AI safety Summit and it’s an attempt to take hold of the narrative and say that this is something that we’re going to be the world leaders on this and we’re going to bring
Together not only other world leaders but important people in Industry like Sam ultman who’s the boss of the company behind chat GPT it’s as much of a PR exercise as much as it is trying to lay the foundations for future regulations or future consensus on AI safety Rishi
Zaka someone who loves technology it’s going to be something that he sees as quite important to his legacy to have hosted this yeah I mean we hear a lot about how the UK wants to become a world leader in Ai and This Global Summit hosting it is is very much part of that
There are a lot of AI startups that are hugely influential in the UK so it’s not necessarily always a case of size but I do think that we’re maybe not as influential in terms of pushing Frontier models forward as businesses in other countries are this is maybe a
Attempt to take hold of the regulation aspect of that and to really be the ones pushing that forward we are a little bit behind the EU in terms of their regulation they’ve already got an act that’s looking to come into Force by the end of the year we do have a white paper
On AI which has given a rough idea what the government stance is and what particular concerns are you hearing that people have raised with you one is about who is represented at The Summit it if you speak to the government they will definitely say we’ve got a big range of
Representation from Civil Society from different parts of the world and from different sizes of companies but it’s a very exclusive invite only event only some hundred or so delegates and it’s been really hard honestly for media to even get to people because of the way that it’s run we have to be escorted
From the media center even to come and record something so we can’t walk around and meet people we can’t gauge the the mood of the people that are here in the way that we’d want to that kind of transparency is lacking a little bit and I would say that what we’re hearing from
People is that there could have been a bit more representation of people from the global South people who maybe are affected by AI bias but are not getting as much say as some of the big tech companies and how to shape our priorities when it comes to Ai and also
More representation from Civil Society people who want to talk about AI ethics and what is happening in terms of this AI supply chain how AI is actually developed and how that can be done better and about transparency and how that works so you mentioned AI bias what
Is that AI is trained on information that we already have in society and so if it reflects Society it will also reflect some of the biases that exist in society whether that is based on gender on race or various other characteristics if we consider some US Government
Research that was published in 2019 and it specifically looked at facial recognition technology which is powered by AI it found that black people were between 10 and 100 times more likely to be misidentified by facial recognition technology than white people and this is a problem because it means that for
Example if it’s used by the police they’re much more likely to say that someone looks like someone else if they’re from a nonwhite ethnicity um and they’ll be over represented among suspects this is something that is affecting people now there have been a number of known wrongful arrests in the
US at least um because of the use of this facial recognition technology and the concern among some is that we focus too much on say the existential risks you know biocurity in 2025 or further in the future when we have these pressing issues at the moment Elon Musk and Rishi
Sunak meeting at Downing Street this afternoon at the time of recording the meeting will have already happened by the time this goes to air is it just a photo op or could it result in some sort of meaningful relationship when it comes to AI it’s clearly mutually beneficial
This partnership for both of them Rishi sunak wants as many eyes on this Summit as possible and he wants it to look like a success and the fact that he’s managed to get one of the big names in Tech and also someone who’s kind of notorious for
Um being online and being known for what he said it’s a good look for him to have Elon Musk on side if we look up from musk’s side he seeks The credibility that being alongside world leaders and being taken seriously gets him but he’s not an AI expert it’s not exactly his
Background yes X formerly known as Twitter uses AI in a lot of its processes but that doesn’t necessarily mean that he’s the person we should be looking to for how we should go forward with AI regulation to better protect people some people will be worried about if it muddies the conversation a little
Bit rather than adds meaningfully to how we’re going to move forward with AI as this first Global AI Summit closes at Bletchley Park this week what’s the feeling there are people encouraged at what’s been discussed this week that AI can be regulated I think there is a sense that people are encouraged by
What’s been discussed but I wouldn’t necessarily say that the focus has been that much on regulation and exactly how it should be done I think if it were you’d see a lot more disagreement so far it’s been more towards getting a consensus on a voluntary agreement about
What we should be doing to ensure AI is safer but what has come out of that has often been more focused on research and um keeping on top of the latest tools and evaluated them for safety there hasn’t really been that much information on what enforcable regulation is going
To look like as a result of this and I think that that is part of the problem because if you focus sure purely on the voluntary agreements you kind of risk allowing big tech companies to mark their own homework because they’ve already developed these models they’ve done so without a great deal of
Transparency we still don’t know what information uh chat GPT has been trained on exactly and now they’re kind of in dialogue with politic to say okay we should be more careful we should be regulating AI we should be protecting people which of course looks great for
Them but we still don’t know how they got to where they are okay artti it’s been great to get your thoughts thanks so much for joining us we’ve talked about AI a lot in past podcasts if you want to scroll back through our episodes and there’s little
Doubt we’ll be talking more about this in the future thanks for your company we’re back again tomorrow
Video Keywords: Artificial Intelligence, SKY NEWS,SKY,NEWS,DAILY PODCAST,PODCAST,AI,ARTIFICIAL INTELLIGENCE,TECH,TECHNOLOGY,BEATLES,NOW AND THEN,BLETCHLEY PARK,ENIGMA,CODE,WW2,WWII,WORLD WAR TWO,SPIES,CODE BREAKERS,SUMMIT,SUNAK,BRITAIN,UK,BRITISH,WORLD ECONOMIC FORUM,WEF,ROBOTS,ECONOMY,JOBS,EMPLOYMENT,IMAGES,POLITICS,SAFETY,ONLINE,CYBER,WEB,CHATGPT,GLOBAL AI COUNCIL,DATA,ANALYTICS,SALLY LOCKWOOD
-
Sale!
Wireless WIFI Repeater Extender Amplifier Booster 300Mbps
$29.99$14.99 Add to cartWireless WIFI Repeater Extender Amplifier Booster 300Mbps
Categories: Electronics, Wi-Fi Router, Wireless Wi-Fi Extender Tags: 300Mbps, 802.11N, Amplifier, Booster, Extender, mobile wi-fi booster, Remote, WIFI, Wireless, Wireless WIFI, Wireless WIFI Repeater, Wireless WIFI Repeater Extender, Wireless WIFI Repeater Extender Amplifier, Wireless WIFI Repeater Extender Amplifier Booster, Wireless WIFI Repeater Extender Amplifier Booster 300Mbps$29.99$14.99 -
Sale!
Full RGB Light Design Gaming Headset Headphones with Mic
$24.99$14.99 Add to cartFull RGB Light Design Gaming Headset Headphones with Mic
Categories: Electronics, Gaming, Gaming Headsets Tags: Design, Full, Full RGB Light Design Gaming Headset, Full RGB Light Design Gaming Headset Headphones, Full RGB Light Design Gaming Headset Headphones with Mic, Gamer, Gaming, Gaming Headset Headphones, gaming headset wireless, Headphone, Headphones, Headset, Light, Mic, Package, RGB$24.99$14.99 -
Sale!
Wireless BlueTooth Multi-Device Keyboard Mouse Combo
$39.99$19.99 Add to cartWireless BlueTooth Multi-Device Keyboard Mouse Combo
Categories: Electronics, Gaming, Gaming Keyboards, Keyboard Mouse Combos Tags: Combo, Keyboard, keyboard mouse combos, Mouse, MultiDevice, Set, WireKeyboard Mouse Combo, Wireless, Wireless BlueTooth Keyboard Mouse Combo, Wireless BlueTooth Keyboard Mouse Combos, Wireless BlueTooth Multi-Device Keyboard Mouse Combo, Wireless BlueTooth Multi-Device Keyboard Mouse Combos$39.99$19.99 -
Sale!
High Back Leather Executive Adjustable Swivel Gaming Chair with Headrest and Lumbar
$199.99$139.99 Add to cartHigh Back Leather Executive Adjustable Swivel Gaming Chair with Headrest and Lumbar
Categories: Gaming, Gaming Chairs Tags: Adjustable, Chair, computer chairs, Desk, Executive, Gaming, Girl, Headrest, High, High Back Leather Executive Adjustable Swivel Gaming Chair, High Back Leather Executive Adjustable Swivel Gaming Chair with Headrest, High Back Leather Executive Adjustable Swivel Gaming Chair with Headrest and Lumbar, High Back Leather Executive Adjustable Swivel Gaming Chairs, Leather, Lumbar, Office, Racing, Swivel$199.99$139.99 -
Sale!
Professional LED Light Wired Gaming Headphones with Noise Cancelling Microphone
$29.99$19.99 Select optionsProfessional LED Light Wired Gaming Headphones with Noise Cancelling Microphone
SKU: N/A Categories: Electronics, Gaming, Gaming Headsets Tags: Cancelling, Gaming, Gaming Headphones with Noise Cancelling Microphone, gaming headset, Headphones, Headset, LED, Light, Mic, Microphone, Noise, Professional, Professional LED Light Wired Gaming Headphones, Professional LED Light Wired Gaming Headphones with Noise Cancelling Microphone, Wired, Wired Gaming Headphones, Wired Gaming Headphones with Noise Cancelling Microphone$29.99$19.99 -
Sale!
Gaming Desk with LED Lights USB Power Outlets and Charging Ports
$349.99$249.99 Select optionsGaming Desk with LED Lights USB Power Outlets and Charging Ports
SKU: N/A Categories: Computer Desk, Gaming, Gaming Desk Tags: and Charging Ports, Charging, Desk, Desks, Gaming, gaming desk with led lights, Gaming Desks with LED Lights, Home, LED, Lights, Monitor, Office, Outlets, Port, Power, Room, Stand, USB, USB Power Outlets, White, Workstation$349.99$249.99 -
Sale!
Wired Mixed Backlit Anti-Ghosting Gaming Keyboard
$99.99$79.99 Add to cartWired Mixed Backlit Anti-Ghosting Gaming Keyboard
Categories: Electronics, Gaming, Gaming Keyboards Tags: Antighosting, Backlit, Blue, brown, Gaming, Gaming Keyboard, gaming keyboards, gaming keyboards and mouse, Keyboard, Laptop, Switch, Wired, Wired Mixed Backlit Anti-Ghosting Gaming Keyboard, Wired Mixed Backlit Anti-Ghosting Gaming Keyboards, Wired Mixed Backlit Gaming Keyboard$99.99$79.99 -
Sale!
Wireless Bluetooth 5.3 ANC Noise Cancellation Hi-Res Over the Ear Headphones Headset
$119.99$59.99 Add to cartWireless Bluetooth 5.3 ANC Noise Cancellation Hi-Res Over the Ear Headphones Headset
Categories: Electronics, Gaming, Gaming Headsets Tags: 5.3 ANC Noise Cancellation Hi-Res Over the Ear Headphones Headset, ANC, Audio, Bluetooth, Cancellation, Ear, Earphone, gaming headset, Headphones, Headset, Hi-Res Over the Ear Headphones Headset, HiRes, Noise, Wireless, Wireless Bluetooth 5.3 ANC Noise Cancellation Hi-Res Headphones, Wireless Bluetooth 5.3 ANC Noise Cancellation Hi-Res Over the Ear Headphones Headset, Wireless Bluetooth 5.3 ANC Noise Cancellation Hi-Res Over the Ear Headphones Headsets$119.99$59.99 -
Sale!
Wired Sports Gaming Headset Earbuds with Microphone
$19.99$9.99 Select optionsWired Sports Gaming Headset Earbuds with Microphone
SKU: N/A Categories: Gaming, Gaming Headsets Tags: Accessories, Earbud, Earphone, Earphones, Gaming, gaming headset with microphone, Headphones, Headset, IOS, Microphone, Sports, Wired, Wired Sports Gaming Headset Earbuds, Wired Sports Gaming Headset Earbuds with Microphone, Wired Sports Headset Earbuds$19.99$9.99 -
Sale!
150W Universal Multi USB Fast Charger 16 Port MAX Charging Station
$49.99$29.99 Add to cart150W Universal Multi USB Fast Charger 16 Port MAX Charging Station
Categories: Charging Stations, Electronics Tags: 150W, 150W Charging Station, 150W Universal Multi USB Charging Station, 150W Universal Multi USB Fast Charger 16 Port MAX Charging Station, 150W Universal Multi USB Fast Charger 16 Port MAX Charging Stations, 150W Universal Multi USB MAX Charging Station, 16 Port MAX Charging Station, 3.5A, Charger, Charging, Fast, laptop charging stations, Max, Multi, Port, Stand, Station, Universal, USB$49.99$29.99
Hopefully AI will be able to debunk all of the wokeness 👍
I'm more worried about islamic takeover
sunak trying to make money just a greedy git
AI will ensure that people become intellectually lazier than they are now.
How risky is it to have corrupt politicians?
this idiot should be focusing on REDUCING TAXES than messing with AI…..another PR stunt…Otherwise we will get Labour in and they will screw things up like that last time!!
Probably no more risky than trusting the BBC & politicians
This man has geek in his DNA. Musk is his god and making money his raison d,etre.
Most poititians have artificial intelligence .
At the Bletchley Park Summit. On the Safety of AI .The unelected 'Prime Minister, said the UK will not" rush to regulate the AI sector. "WHY ? Well one of Old Rustys Favourite business chums. Mark Andereessen a partner of VC Firm Andreessen Horowitz wants to open up a corner shop in London. Who "shock Horror " is one of the world's most top investors in AI business Trebles all around. So old Rusty Fiddles while people lose (eventually) there Jobs
It's still not even close. Still just a tower of IF THEN statements. We are yet to get anything actually thinking..
If Sunak and Musk are in charge then humanity is doomed.
"Sally" sounds like AI.
I am glad i am getting older and hopefully i wont be around for AI. As Stephen Hawkings said AI could be the end of humans. I dont doubt they could benefit us but there are so many evil nutters who will abuse it or they will out smart us. We cant control ourselves never mind AI
Very Risky – I'm not sure which is worse. Artificial Intelligence or natural stupidity.
Has the P M sorted out making Remembrance Sunday safe from the violent islamic / palasinian nut jobs , so we natives can go to the servcers in peace ? ? ?
AI is likely to be sold as a great blessing to us . I suspect that , ultimately , it will be used to replace and police us . I sense this is the plan of a few powerful individuals and corporations . People will find out too late that they are no longer needed and that their jobs are taken by super intelligent machines . The media is going to go along with the " no threat " narrative – until it too is replaced . Expect disaster !
It can't be any worse that being led by the current globalist politicians. They seem intent on destroying cultures, history and national pride right across the western world.
So risky proper comment deleted, even Gods words, by the delete master
I assume Sunak has investments in AI companies
All IT projects are risky, mainly because they dont work.
Trying for a long time to remove emotion but now trying to find rage! Lost again
Sky censoring comments about how this is another Rocky Fella scenario.
Hak attak
The 'AI' threat comes out of the same Rockafella playbook that the 'Lockstep' pandemic scenario did.
Look up 'Scenarios for the Future of Tech and International Development' by the Rockyfella Foundation. This latest scenario is called 'Hack Attack'.
The 'AI' threat comes out of the same Rockefeller playbook that the 'Lockstep' pandemic scenario did.
Look up 'Scenarios for the Future of Technology and International Development' by the Rockefeller Foundation. This latest scenario is called 'Hack Attack'.
'Daily podcast: How Rishi is artificial intelligence?'
Ask joe
How risky is it having lefty idiots running all media organisations