The AI series with Maria Ressa: An introduction | Studio B: Unscripted

Summary

Journalist Maria Ressa interviews AI expert Mike Widge on the transformative impact of artificial intelligence. They discuss AI's rapid growth, potential risks including misinformation and cyber threats, and the importance of guardrails. Widge highlights AI's role in exacerbating societal issues and the need for regulation to prevent misuse, emphasizing the urgency of addressing AI's societal effects.

Takeaways

• 🌐 Artificial Intelligence (AI) is transforming various aspects of human life, from work to warfare, with both opportunities and risks.
• 📈 The progress in AI accelerated significantly after 2022, driven by advancements in machine learning and the availability of big data.
• 💾 AI relies heavily on data; social media posts, for example, contribute to training AI models by providing labeled data.
• 🔍 AlexNet in 2012 marked a turning point in AI capabilities, demonstrating a leap in image analysis and interpretation.
• 🧠 Large language models like ChatGPT function on a principle similar to smartphone autocomplete, but at an unprecedented scale.
• 🌐 The training data for AI models is vast, with the entire worldwide web as a starting point, including only 3% from Wikipedia.
• 🛠️ AI's potential dystopian outcomes range from job displacement to existential threats, though current consensus suggests AI will augment rather than replace human jobs.
• 🌿 There's optimism that AI could help solve major global issues like climate change, with applications in fields such as synthetic biology.
• 🚩 The misuse of AI poses significant risks, including cyber-attacks and information warfare, which are more pressing concerns than sci-fi scenarios.
• 🔒 Transparency and guardrails are crucial for responsible AI development, to prevent unintended consequences and misuse.
• 🗳️ AI's impact on democracy and electoral integrity is a pressing concern, with the potential to industrialize disinformation on a massive scale.

Q & A

What is the significance of the year 2022 in the context of AI development?

The year 2022 is significant because it marked the launch of Chat GPT, which led to exponential growth in AI capabilities and interest from big tech companies.

What is the role of data in training AI systems?

Data is essential for training AI systems. It is used to train neural networks, with social media uploads often serving as training data.

What was the impact of AlexNet on AI development?

AlexNet was a pivotal AI program that demonstrated a significant leap in image analysis capabilities, marking the beginning of a new era in AI.

How does a large language model like Chat GPT function?

Large language models like Chat GPT function by predicting the most likely completion of a text input, similar to smartphone autocomplete features, but on a much larger scale.

What is the scale of data used to train large language models?

The scale of data used is immense, starting with downloading the entire worldwide web, with Wikipedia constituting only 3% of the training data.

What is the best-case scenario for AI according to the discussion?

The best-case scenario for AI is that it becomes a tool used by most people in their jobs, enhancing productivity without replacing human roles.

What are the existential risks associated with AI?

Existential risks associated with AI include the potential for AI to become so powerful that it could lead to the end of humanity if it can self-improve without human supervision.

How can AI impact democracy and societal structures?

AI can impact democracy and societal structures by enabling the spread of misinformation and manipulation, influencing public opinion and potentially undermining electoral integrity.

What are guard rails in the context of AI?

Guard rails in AI refer to the safety measures and protocols implemented to prevent AI from generating inappropriate content or being used maliciously.

Why is transparency in AI development important?

Transparency in AI development is important to understand the training data used and to ensure that AI is not inadvertently promoting harmful biases or behaviors.

How can individuals protect themselves from the negative impacts of AI?

Individuals can protect themselves by being aware of AI's potential to manipulate, seeking information from trusted sources, and understanding how their data is used.

What is the future of jobs in relation to AI?

AI is expected to change the nature of work, with some tasks being automated, but it is unlikely to replace all human jobs, especially those requiring creativity and empathy.

Outlines

00:00

🌐 Introduction to AI's Impact

The paragraph introduces the transformative power of artificial intelligence (AI) and its profound potential to change human history. It mentions AI's influence on various aspects of life, from work to warfare to societal structures. The speaker, Maria Ressa, a journalist and Nobel Peace Prize laureate, discusses the dual nature of AI as both an opportunity and a risk, highlighting her personal experience with online harassment and the role of social media in spreading misinformation. She emphasizes the importance of understanding AI, setting the stage for a discussion with Professor Mike Wridge, an AI researcher with over 35 years of experience.

05:01

📈 AI's Evolution and Data Dependence

This section delves into the history and development of AI, noting that despite its seemingly recent surge in prominence, AI has been an area of study since the 1950s. The slow progress in AI is attributed to the lack of computational power and data until this century. The advent of big data and increased computing capabilities have been pivotal in propelling AI forward. The discussion highlights how social media contributes to AI training by providing vast amounts of data through user interactions and content sharing. The paragraph also introduces AlexNet, a pivotal AI program in image analysis that marked a significant leap in AI capabilities.

10:02

🧠 Understanding Large Language Models

The conversation explains the workings of large language models like Chat GPT by drawing an analogy with smartphone autocomplete features. These models are trained on extensive datasets, including the entire worldwide web, to predict and generate text based on patterns learned from the data. The scale of data used is immense, with Wikipedia constituting only a small fraction of the total training data. The potential applications of AI are broad, ranging from solving complex problems like climate change to concerns about AI's role in replacing human jobs, which the speaker refutes as unlikely.

15:04

🚀 AI's Role in Society and Misinformation

This paragraph addresses the dichotomy in public perception of AI, swinging between utopian and dystopian views. It emphasizes the need for a balanced understanding and identifies realistic concerns such as AI's potential to enable malicious activities and cybersecurity threats. The discussion also touches on AI in warfare, the lack of regulations, and the ethical implications of AI's unchecked advancement. The potential for AI to be used in information warfare, influencing public sentiment and electoral outcomes, is highlighted as a pressing issue.

20:04

🛡️ Guardrails and AI's Unintended Consequences

The dialogue focuses on the importance of implementing 'guardrails' or safety measures in AI to prevent unintended harmful consequences. It acknowledges the current measures as inadequate and the need for transparency in AI development. The conversation raises concerns about AI being developed by a few entities with potential biases in training data, affecting society negatively. The rapid evolution of AI and its impact on democracy, mental health, and the potential for personalized misinformation campaigns are also discussed.

25:06

🏛️ AI in Warfare and the Future of Work

This section discusses the international community's stance on lethal autonomous weapons and the ethical concerns surrounding AI in warfare. It also addresses the use of AI in recruitment and the potential for AI to replace human decision-making in the workplace. The speaker advocates for human involvement in critical decisions and expresses concern over the dehumanization of jobs through AI. The paragraph concludes with a contemplation on the future of jobs and how AI might alter traditional work roles.

🌟 Final Thoughts on AI

In the concluding paragraph, the discussion is summarized with a call for caution and responsibility in the development and use of AI. The speaker emphasizes the need to balance technological advancement with ethical considerations and societal impact. There's an acknowledgment of AI's potential to bring about significant changes in how future generations live and work, while also stressing the importance of maintaining human values and relationships amidst technological progress.

Keywords

💡Artificial Intelligence (AI)

Artificial Intelligence refers to the simulation of human intelligence in machines that are programmed to think like humans and mimic their actions. In the video, AI is discussed as a transformative technology affecting various aspects of life, from work to warfare. The script mentions AI's potential to bring profound changes and its role in social media, highlighting both opportunities and risks.

💡Existential Risk

Existential risk is the risk of an event causing the extinction of humanity or a significant fraction thereof. The video discusses AI as a potential existential risk, particularly if it becomes uncontrollable or misused, leading to scenarios that could threaten human existence.

💡Disinformation Campaign

A disinformation campaign is a deliberate effort to spread false information to influence public opinion or obscure the truth. The script mentions Maria Ressa's experience with such a campaign, illustrating how AI can be weaponized to spread lies and hate, undermining democracy.

💡Social Media

Social media refers to digital platforms such as Facebook, Twitter, and Instagram, where users create and share content. The video discusses how social media platforms prioritize the spread of provocative content, which can be exploited by AI to disseminate disinformation.

💡Large Language Models

Large language models are AI systems trained on vast amounts of text data to generate human-like text. The video explains how models like Chat GPT work by predicting text based on patterns learned from data, raising concerns about the quality and impact of the information they generate.

💡Data

Data refers to the information, often in digital form, that AI systems use to learn and make predictions. The video emphasizes the importance of data in training AI, noting that activities like social media posting contribute to the data that AI systems use.

💡Neural Networks

Neural networks are computing systems inspired by the human brain designed to recognize patterns. The video mentions that AI requires large neural networks, which are built with significant computational power and data, to function effectively.

💡AlexNet

AlexNet is a significant AI program that marked a leap in image recognition capabilities. The video uses AlexNet as an example of a milestone in AI development, illustrating the rapid advancements in the field post-2012.

💡The Singularity

The Singularity refers to a hypothetical point in the future when AI becomes smarter than humans and capable of self-improvement. The video discusses this concept as a potential existential risk, where AI could surpass human intelligence and control.

💡Cybersecurity

Cybersecurity is the practice of protecting systems, networks, and data from digital attacks. The video mentions cybersecurity as a significant risk amplified by AI, where bad actors could use AI to launch sophisticated attacks.

💡Electoral Integrity

Electoral integrity refers to the fairness and accuracy of elections. The video discusses how AI could be used to disrupt electoral integrity through the spread of misinformation, influencing public opinion and election outcomes.

Highlights

Artificial intelligence is transforming our world and expected to bring profound changes.

AI's impact on work, warfare, and societal structures is significant.

AI presents both huge opportunities and risks, including potential existential threats.

Journalist Maria Ressa discusses the dangers of AI and its threat to democracy.

Social media design prioritizes the spread of lies and anger, exacerbated by AI.

AI has been studied since the 1950s, but progress accelerated in the 21st century with increased computing power and data.

Data is crucial for training AI, with social media providing a wealth of training data.

AlexNet in 2012 marked a significant leap in AI's image analysis capabilities.

Large language models like ChatGPT work on a principle similar to smartphone autocomplete.

AI training involves downloading the entire worldwide web, with Wikipedia constituting only 3% of the data.

AI's potential to solve major problems like climate change is discussed.

AI is unlikely to replace human jobs entirely, but will become a tool used in jobs.

The concept of 'The Singularity', where AI surpasses human intelligence, is explored.

Current AI risks include cyber attacks and AI weapons of war, rather than existential threats.

The importance of guard rails and checks in AI to prevent the spread of inappropriate content.

AI's impact on society is a concern, with the technology evolving faster than government regulation.

The potential for AI to be used in disinformation campaigns, impacting electoral integrity.

AI's role in the job market and the concern about it filtering CVs and introducing biases.

The future of AI in warfare, with discussions on lethal autonomous weapons.

AI's potential to change administrative jobs, making human involvement seem strange in the future.

Transcripts

[Music]

artificial intelligence AI is already

Transforming Our World and is expected

to bring some of the most profound

changes in human history think of me as

a friendly companion who can provide

helpful insights from the way we work to

the way Wars are fought to the very

fabric of our societies it seems set to

bring huge opportunities but others warn

it could lead to our own destruction

it's one of the existential risks that

uh we facing and potentially the most

pressing one my name is Maria oressa and

I'm a journalist from the Philippines

through our investigations I became the

target of a harassment and

disinformation campaign receiving

thousands of death threats online I

received the Nobel Peace Prize in 2021

an acknowledgement of how difficult it

is for journalists to do our jobs today

I saw firsthand the dangers of tech and

its threat to to democracy the design of

the systems of social media prioritizes

the spread of Lies laced with anger and

hate in this special series of Studio B

on artificial intelligence I'll be

meeting some of the brightest Minds

working in the field today my guest this

week is Professor Mike wridge he's been

working in AI research for over 35 years

in Oxford and at the prestigious Allan

touring Institute a prolific author he's

written nine books and over 4 400

scientific articles on the subject so

what exactly is artificial intelligence

how did we get here and is it really a

threat to our very

[Music]

existence

Mike it is so good to see you and you

know you have been studying artificial

intelligence for 35 years but something

changed right it grew exponential

exponential right after November 2022

after chat GPT was

launched how did we get to where we are

first how do you define where we are

what does the science tell

us so artificial intelligence despite

appearances is not a new field it's been

studied very very actively since uh

since since the 1950s but the truth is

that actually progress in AI was really

glacially slow until this Century um

computers in the past just weren't

powerful enough and we didn't have the

data and we are in the world of big data

and AI is nothing without data you

absolutely need data to to to train AI

to use the terminology and every time

you upload a picture of yourself to

social media and you helpfully label it

with your name or your kids do what they

are doing is providing training data

social social media companies that's

literally what their role is in in doing

that so you need data and you need lots

and lots of computer power to be able to

build neural networks that were big

enough so around about 2012 or so Alex

tell us about Alex net so Alex net um

was a computer program and AI program to

do basically image analysis and it was

entered into a competition and entries

in this competition were judged at how

well they could interpret pictures in in

images and the point about Alex net was

that in one year we saw a step change in

capability and this got everybody's

attention and it it became clear at that

point that we really were in a kind of a

new era of AI and that was the point I

have to say that the big tech companies

noticed uh and started to get really

really really interested can I ask you

something very geeky

uh you talked about training data about

machine learning artificial intelligence

neural networks large language models

how do these all fit together okay the

way that large language models like chat

GPT work is really bizarrely simple it's

just doing exactly what your smartphone

does when you do autocomplete so if you

open up your smartphone uh and you start

sending a text message so for example I

start sending a tech message to my kids

and I type have you it will suggest

completions and the completions might be

tided your room or walked the dog right

those might be the likeliest completions

of that so how is it doing that it's

been trained on all of the text messages

I've sent my kids and learned that the

likeliest completions of have you are

either going to be walked your dog or or

tidied your room so chat GPT is doing

nothing more than that the difference is

the scale right chat GPT is built with

AI supercomputers that run for months uh

and cost tens of millions of dollars to

be able to do that training and the

training data is basically it's not your

smartphone messages it's all the Digital

Data available in the world and the

standard way that you build these is to

start by downloading the whole of the

worldwide web right the entirety of the

worldwide web so Wikipedia makes up just

3% of the training data for uh for these

large language models so the scale of

the data is incredible um there are

people who say uh that this will solve

humanities worst problems like climate

change deep mind uh which is behind

Google Search now because they bought it

also does synthetic biology right and

that maybe we can use phytol plantant

that can pull carbon out of the air like

can you give me your best case

scenario okay so it is quite remarkable

that the discussion around AI either

veers to the extremely dystopian it's

going to be the end of humanity or the

extremely utopian and there's not

actually a lot between those two the

reality is going to be between those two

the idea I mean I think Elon Musk was on

record recently as suggesting that AI

was going to take all our jobs that

seems very unlikely to me not in the

lifetime of anybody in this room AI will

become a tool that most people use in

their jobs but it's not going to replace

people I mean for example there are

going to be lots of applications of AI

in education which is going to be really

wonderful but what teachers do is a very

human thing it's not going to replace

all of humanity and allow us to spend

our lives writing poetry or whatever it

is that we would do if we did didn't

have uh jobs I think so that scenario is

is extremely unlikely the dystopian

scenarios have been really hotly

discussed and people talk about

existential risk and that literally

means the end of humanity that AI could

become so powerful that somehow it ends

Humanity if it can program itself if it

can get resources it can continue doing

it without human supervision right so so

there's this scenario uh called The

Singularity yes and it's a beautiful

scenario which makes for great science

fiction and the idea is only fiction go

go the idea is at some point in the

future at some point we don't know when

AI is going to be as smart as we are and

at that point it can start to improve

itself it can literally rewrite its PR

code and then at that point it's smarter

than we are and that improved AI can

then improve its code again itself uh

and it just continues that process and

the fear is at that point that AI is out

of our control I saw this on Black

Mirror and actually of all of the

Contemporary science fiction shows Black

Mirror I think is absolutely by far the

best it's very thought-provoking stuff

so is it so i' in all of this discussion

I've never seen a single genuinely

plausible scenario for existential

threat and it really has been discussed

endlessly with some very very very smart

people thinking about it the biggest

risks right now are that AI is a

powerful tool yeah and it enables bad

people to do bad things bad things that

they couldn't previously have done that

it enables a whole category of risks not

existential risks but risks like cyber

security attacks which would just not

have been feasible uh previously that I

think focusing our attention on those

issues I think would be much more

productive than on science fiction

issues or AI weapons of war right like

AI drones which they've used in Ukraine

are being used in Moscow um there again

are no boundaries set on this and yet

the

scientists with a profit motive are

rushing ahead and we are like Pavlov's

dogs in real time how can we protect

ourselves in this because if you look in

in the Nobel lecture in 2021 I actually

said that um we had data that showed

that we're being insidiously manipulated

it has so much of our data

that it cuts in through our emotions

information Warfare changes the way we

feel changes the way we think and then

the way we act electoral Integrity for

example I mean I don't I don't think

it's a coincidence that you have now

according to VM 72% of the world under

authoritarian rule right so these are

some of the impact of it Mr scientist

tell me right because you don't have a

profit motive right you're studying the

science how do we rein them in

so I think there's there's two sets of

issues the first is we we're pretty

confident right now that social media

one of the unintended consequences of

social media was a mental health crisis

in teenagers and we didn't see that

coming right but that's just one of the

unintended consequences and I think what

you're saying is what are the unintended

consequences of AI going to be so for

example what if we end up with some

future large language model which just

completely inadvertently makes us more

aggressive

or more depressive for example and what

impact would that have globally for

example a widely used AI tool that made

us more aggressive might lead to more

conflict in but isn't that happening

since they took all of the Big Data the

unstructured Big Data of social media

full of fear anger hate right isn't that

happening now okay as we already

mentioned the way that this technology

is configured is you download the whole

of the worldwide web now you don't have

to look very hard on the worldwide web

to find all sorts of unpleasant and I

mean if you go on you know some social

media platforms they have types of

unpleasantness that we could scarcely

imagine right so and if all of that has

been absorbed by a large language model

then it's a seething cauldron of

unpleasantness now I think genuinely you

know responsible AI companies have no

intention whatsoever of uh unleashing

that on the world so what they do is

they're building guard rails and so they

try to intercept queries that are how do

I build a pipe bomb they will try to

inter such a and also they will look at

the outputs of the large language model

and try to intercept which is

inadvertently com out with which is in

appropriate at the moment those guard

rails I think are the technological

equivalent of gaffa tape they're just

being you know they're being plastered

onto these exactly there's no deep

fixers to that and one of the worries is

if this technology is owned by a small

group of actors who develop this

technology behind closed doors we don't

to see the training data so you have no

idea what this has been trained on about

you and you're a public figure there

would have been a great deal of content

about you and some of it won't have been

very nice that's a safe bet so this is I

think a real concern and this issue of

transparency I think is is is really a

concern which needs to be taken very

very Ser you talked about guard rails

right there's no incentive for them to

put guard rails in I mean they the only

incentive is that they won't be attacked

by people right it's a reputational

thing but if they can get by without it

they have as they have with social media

we still haven't done anything well I

think here we are in a situation uh

which is very awkward we've got AI which

has gone viral uh it's the first large

language models are the first general

purpose and I'm choosing those words

very carefully general purpose AI tools

that have reached a mass market and

they're very powerful yeah and the tech

companies see Empires they see na

Empires and they want to they want to

stake their claim on those Empires they

want to be those Empires they want to be

the Google they want to be the they want

to be the Amazon of the generative AI

world and the very big risk is that what

they're doing to try to get an advantage

on their competitors is Rush ahead with

this technology without thinking about

for example whether it's really fit for

prime time and that really is a worry um

but these worries are not you know

they're not unknown I mean the the UK

government convened an international AI

safety Summit and I have to tell you

there was a there was some skepticism

about what it was going to achieve but

actually the the debate was was a

sensible debate uh and it got it on the

international agenda so I think what's

going to be challenging is the extent to

which government can really hold the

richest companies in the world to

account and the irony of course is that

if they get it out to all of us they get

all of our data we train their large

language models they gain more power

even as the very nation state that are

going to try to put regulations in place

to control them lose power because the

technology is already impacting Society

all around the world right it's a tough

one um I guess you know in I I have a

bleak picture of this as you can tell

because having been

attacked and to hear them say you know

well we didn't intend that doesn't

really matter what the intent was and so

how do we what can we do right now right

it's moving too slowly governments move

at the pace of years while the tech

evolves in every two weeks Agile

development means they're rolling out

code every two weeks right so is there

anything anyone watching can do well I

think there is there is concretely

something we can do so we're heading

into elections in the UK and the US the

US India one of the very prominent risks

with this technology is the possibility

of industrializing the production of

disinformation and misinformation uh uh

on a massive scale on an unprecedented

scale and personalizing it down to the

level of individuals so the AI can look

at my social media feed and pick up on

the sentiments that I express in my

social media feed pick up on my

political stance which is going to be

implicit within sometimes explicit

within my social media feed and then

feed me personally tailored very high

quality misinformation the sentiment

analysis exists to do that the

generative AI makes it possible to do

that and uh and and the cost of

launching a disinformation campaign in

an election because of generative AIS

come down massively yeah and let's be

honest there are people in the world

with a huge interest in disrupting

elections in the US or the UK or India

and so on uh and it could be people just

with an interest in vandalizing the

process or it could be state level

actors that really want to disrupt

what's going on so what can we do

concretely I think we absolutely need to

be alert to that issue I think trusted

news sources are going to become so

valuable the difficulty with that of

course is that we end up in a world

where you know we're all completely

paranoid and don't believe anything but

trusted news sources I think are going

to be essential and understanding how we

can be manipulated I think is really

really important there is so much more

we can talk about this because in 2024

one in three people around the world are

going to vote and this is the Tipping

Point for both electoral systems our

democracies um but we're getting to the

Q&A so let me toss it to you the

gentleman in the back was the first hand

up leading on from what you were saying

Michael about news and trusted news

given the growth of generational AI

technology which can actually do deep

fakes pretty convincingly both in audio

and video uh on social media feeds how

long before we the poor public cannot

tell the difference anymore

well I think uh in terms of being able

to tell the difference AI right now can

perfectly duplicate your voice to the

point where nobody would be able to tell

the difference that's a technology which

exists deep fake images uh is not quite

there but very very close but I don't

know if people saw do you remember

seeing this picture of the Pope in this

big puffer jacket that went viral and I

have to tell you when I first saw that I

didn't actually twig that this was not a

a real image I just assumed it was and I

thought this was a a slightly strange

clothing choice for the pope um so we

need to raise our guards for that and we

need I say the the the issue of trusted

news sources is just going to be so so

so important you know they're going to

be facing this technology and they're

going to need to think of new ways of

dealing with that technology but let's

hope they can rise to the occasion I

mean I I'll pick it up and I'm slightly

more

pessimistic um I promise you won't walk

out depressed completely right um I

think our shared reality is already

broken the political Dominoes on social

media of information operations the

political dominoes fell in 2016 duterte

was elected in the Philippines in May

about a month later you had brexit and

then you had all of the elections moving

Trump was elected in November and you

know we have the data to show that there

were information operations that were

there it plays to our fears our hatred I

mean this is playing out right now and

our shared reality is splintered right

so what do we we have to do um news

organizations are under attack both on

the business model side right um the

money that used to go to news uh we

still have to maintain very expensive

systems of checking everything because

we stand behind it we're legally liable

that now goes to microt targeting microt

targeting is not the same as advertising

it goes to your weakest moment to a

message that's and it's cheap right so

that's still there and then Mike is

going to bring large language models we

have it there and we've already seen

this wow I sound really Bleak it's only

cuz I was getting 90 hate messages per

hour and in order to keep doing my job I

had to be okay with going to jail for

the rest of my life that's a lot to ask

from your journalists so what do we do

come out in the real world right

understand you are being manipulated up

until the guard rails are put in place

we need to organize ourselves and have a

shared reality this is a shared reality

right now so we want to try to get more

questions from the audience go ahead

saadya UK campaign to stop Killer Robots

um and my question is um in your view

what do you think the International

Community should be doing uh to address

the concerns and challenges of the use

of AI in Warfare given that we're seeing

this being used in Ukraine and in isra

Gaza well again we're stepping well

outside my comfort zone I can tell you

firstly the international AI Community

is broadly but not universally against

lethal aonomus weapons so in 2015 I was

organizing a conference and we had a

panel on exactly this topic and I

thought the the views were going to be

absolutely unanimously against lethal

autonomous weapons and was really

startled to discover that there are

people of good faith who think that no

that this is this would this is how my

children can avoid having to be involved

in warfare that's literally how some

people viewed this so it was a it was a

more complex issue than I thought but I

I can tell you what my perspective is

and my perspective is that I do not

think it's acceptable that a machine

decides autonomously whether to take a

human life if a human life is taken then

which uh you know is is uh is is an

extremely undesirable situation in any

case but somebody who takes that

decision on a battlefield has to be able

has to be capable of empathy and

understand the consequences what it

means for a human being to be deprived

of their life what can we do with it

well we've we've moved on landmines

internationally imperfectly but that

shows that there are ways ahead with

this at the same time we need to be

realistic there are nation states that

are not remotely interested in the

niceties of these issues and they will

develop this technology uh in secret and

we won't see that and so we do have an

obligation to make sure that we can

protect against the the attacks by that

kind of technology I think that's the

only realistic and respons responsible

Way Forward but um by and large the the

Crum of comfort I can offer you is that

the vast majority of the AI Community

think the technology is AB Boren and

want nothing to do with it let's take

one last question uh so AI has been used

more and more in recruitment uh to stiff

through vast amounts of CVS which for

people of my age is a bit of a concern

and obviously a lot of other people as

well uh the sad thing about this is that

a lot of people don't actually know this

and also don't understand what language

it picks up into text and it doesn't

necessarily look at niches so is this a

concern that people might miss good CVS

and people that would actually be the

perfect candidate because they don't

know this and what can be done to

eliminate this or at least reduce the

risk in the future or the inherent

biases in the technology actually make

those selections I I'm completely with

you on this one I don't think this is a

good use of the technology I want humans

involved in that decision this to me is

just lazy and inappropriate HR practice

if a HR manager says oh but the

decisions will be fairer I think it's

absolutely nonsense I just don't think

that that's something that we should do

but if you think that's not a nice use

of the technology imagine you've got AI

as a boss telling you what to do Moment

by moment through your working life and

there are companies pursuing that you

know looking at every email that you

send commenting on it you know wridge

you know you only sent 20 emails today

the company average is 22nd you took

five bathroom breaks today the company

average is three that kind of do you

want to live in a world where AI is

giving you that kind of people I don't

and I dare say you don't either there

are companies pursuing that nonsense now

yeah yeah we won't name it okay we we do

get one more question so let's end with

the gentleman in the back he's hand up

first I have a very simple question um

we talk about loss of jobs due to AI um

do we think in a hundred years time it

will be strange to explain to a child

that um the jobs being done in admin and

companies would have done by humans I

think they will find that very

strange well the f future is going to be

not just weirder than we imagine but

weirder than we can imagine I have

teenage kids who've grown up with the

internet and they just assume that it's

there and it's always on and when it

doesn't work for whatever reason they're

just perplexed and don't you know

something's gone wrong with the world if

the the internet doesn't work for them

kids that are seven or eight years old

now are going to grow up they're the

first generation in history that's going

to grow up being surrounded by very

powerful general purpose AI tools like

chat G PT and they are going to do the

weirdest things with it and the the best

example I can I can give you is is you

go back to the origins of YouTube which

is 2005 or so and you nobody really knew

what it was at the time you could upload

family videos and share them with family

members or you know you could upload

clips of your favorite TV shows nobody

predicted YouTube influencers or the

fact that people would not just be able

to make a living but actually make a

fortune by making videos of themselves

playing computer games and talking over

it and people do you know nobody

predicted that in exactly the same way

no we can't predict right now how our

kids are going to use uh AI in the

future but the basics of humanity are

not going to change they didn't change

with rock and roll they didn't change

with television they didn't change with

Cinema they didn't change with novels

the fundamentals of humanity and human

relationships are going to be the same

but our kids are going to be creative in

ways that we just find I say weird and

hard to imagine but for them it's going

to be it's going to be a

ride he's again very optimistic it is a

ride right but I think this is this

moment in time uh this moment in time is

critical we need the science um and we

need to curtail the for-profit motive so

that we can be safe with the technology

absolutely Michael wrid thank you thank

you so much for joining us in Studio B

the AI series I'm maressa thank you for

joining

[Applause]

[Music]

us