Sign in. It’s quick, free and it’s up to you.
An account is an optional way to support the work we do. Find out more.
WE ARE AT the start of some of the most significant changes in human history that will result from the increased use of artificial intelligence (AI) in virtually every area of life. The speed of some of these changes will be exciting or frightening, depending on your perspective.
In medicine, AI is being used already to diagnose and remotely treat patients, as well as to identify potential new drugs and make prescriptions. Job applicants are shortlisted using AI and credit ratings are determined by the same technology. Data is increasingly being used to determine large swathes of public policy, from policing to transport planning.
In science fiction, the cyborg is this part human, part machine where the technology can carry out many of the traditional functions at speed of the human. If we consider all of the actions and decisions that our mobile phones now carry out on our behalf, could we already be considered as cyborgs?
Rapid developments
In the 2030s, driverless vehicles will mean significant changes in employment of those traditionally behind the wheel, while the advance of 3D printing combined with AI planning and design will see buildings constructed in unprecedented ways.
AI will be able to detect patterns in weather patterns and in markets faster than meteorologists or economists or traders. This can be advantageous if applied to the good, protecting during adverse weather or ensuring that vital product supply chains remain open. But greed could also see such knowledge derived from AI used to advantage the few.
Our lives are arguably being made easier with the use of AI but there continue to be serious concerns. The algorithms that AI uses are based on the data that they are fed. There is always a need to safeguard against particular bias in that data. There are well-documented cases of algorithmic bias in some instances against women, people of colour, certain minorities.
There are obvious concerns about the extent to which machines can make decisions. The weaponisation of AI by States or by private parties is a key debate for multilateral organisations.
Where are the policies?
We, therefore, need to be prepared from a policy and planning perspective but also to adapt our education system to one where engagement with AI is at its centre. These issues will be at the centre of political and philosophical debate over the next decade.
In July 2021, the Government published a national strategy on AI – ‘Here for Good’. There was a strong emphasis on building public trust in new technology and that its use should be ethical and people centred.
The European Union is currently moving toward adopting the Artificial Intelligence Act, in a move that will regulate the increased use of automated decision making by public and private institutions. Much like how the General Data Protection Regulation (GDPR) set the bar around data privacy and protection, the EU is hoping that the new Act will set the standards around how we use AI.
We need to regulate and monitor how AI is used in public and private hands. The risks of a surveillance society, such as that in China where every citizen is monitored by the State and given a credit score, are very real. The increased power being subsumed into the boardrooms of a small number of global tech companies also needs to be checked to ensure that AI is not misused.
Given the speeds at which AI will be able to operate, there is a case for global agreement on where human intervention is appropriate. The risk of cyberwarfare is very real. We need to look at mechanisms to prevent automated decision making that could lead to destruction.
Ethical challenges
There will be deep ethical questions. Should we allow AI to recreate avatars of those who have died? In making judicial decisions, what role should AI play? For us as legislators, in the impossible battle to keep pace with trying to regulate new technologies, to what extent might we have to use AI to guide us?
Artificial intelligence will be transformative and our education systems will have to adapt. Very soon, students will simply don virtual or augmented reality headsets and be educated immersively with the aid of machine learning. But there will be broader questions. In the same way that debates happen around children’s exposure to screentime, families and society will debate whether children can effectively be minded by AI babysitters or befriended by humanlike robots.
Sectors of society will spend the majority of their time in an online metaverse while others will reject new technologies and choose to live, Amish-like, in an AI free world. How do we build a social contract when society may divide in such a way?
Of course, AI (thus far!) lacks those very qualities that should influence all of human activity – emotions such as love, compassion, fear – even though increasingly machines are learning to recognise such expressions. We need to ensure that those qualities that make us human continue to guide our decision making.
The debate about our relationship with Artificial Intelligence cannot be confined to a small minority, mostly within the tech sector. This is a critical debate that demands the attention of politicians, policymakers and the public at large. It will become one of the dominant themes of public discourse over the next decade.
Malcolm Byrne is a Fianna Fáil Senator and the party’s spokesperson on Further & Higher Education, Research, Innovation & Science.
To embed this post, copy the code below on your site
This government trying to discuss any form of intelligence seems like an oxymoron of sorts.
There’s a really good joke in here somewhere.
@Fergal McDonagh: so what are they supposed to do…. Sit down and say nothing… Grow up. Your dislike for the government is OK, but they are working on things you and I don’t fully understand. Intelligence is something that can be viewed as being pompous and stupid by the threatened
@Peter Roche: please do not insult the rest of us if you think you “don’t fully understand” something a politician is working on.
If AI takes over in all aspects such as medicine, engineering, driving, maybe in construction in time, where will the jobs be? Universal income is a means of communism if nobody works over AI efficiency. There are positives but this could be a runaway train too
@Kieran: That hits one of all too many nails square on the head. Every society and economy to date is based on scarcity we have no idea how to live without it but we’re probably about to have to learn. It’s not too hard to envision good endings but the route from here to there is far from clear and looks to be through several minefields.
Is it communism if there are no workers?
@Kieran: Two/three day work week. If AI takes over every service & product, then everything trends toward being free. The rich and powerful may act in unpredictable ways as they lose dominance. Even maintaining control of rare materials is undone by the limitless resources in the asteroid belt.
For the foreseeable future I think, artificial intelligence will remain artificial in the sense of being simulated and not ‘real’ – there will be no-one home even if it seems that way – see the recent Blake Lemoine story on Google’s LaMDA research chatbot: he thought the bot was sentient as it can riff on the topic of sentience, life and death, but it could equally riff on the question of what it’s like to be a vanilla ice cream camel if prompted. It’s a very complex language model but no more than that.
The number of neurons and the way they are layered will simply be too small for now and presumably years to come. We have on the order of 80-90 billion neurons, which about the number of stars in our galaxy. It’s hard to get to even a fraction of that with artificial deep-learning neural networks, but more importantly we’ve little idea how to approach the connectivity to achieve sentience.
In the periphery of the brain our neural networks tend to be about perception – hearing, seeing, tactile feeling – but as you go deeper the networks get more abstract and high-level; you start asking questions about what you’re seeing, and deeper down you may raise doubts, and out of all of those billions of neurons together, but only if wired and trained in the right way, will we eventually begin to see something emerging that may be like a consciousness. I do not think this is impossible to achieve artificially in principle, but I do think we are a long way off.
In the meantime we’re going to see people interacting with (corporate) AI bots, and they’re going to be thinking there’s somebody home while there isn’t. These bots are trained on inputs to generate ‘correct’ outputs, but there won’t be any creativity – it’ll be merely summary outputs of whatever the data dictates.
@Mick Tobin: Did you see the new MidJourney AI that generates images from keyword inputs?
https://www.reddit.com/gallery/x381r2
https://www.reddit.com/gallery/x12nqz
What does this mean regarding creativity?
I understand the AI is trained on billions of images, but but the images it creates aren’t recreations of exiting images, it generates new unique images, a representation / impression of what humans created. It might not be strictly art, but it’s at least creative.
@Mick Tobin: I think he is referring to the ethical conscious of the humans rather than the machines. Where do we draw the lines of what is too far. For example what I know of this metaverse society I have concerns it will damage the social and mental wellbeing of many that use it. But perhaps the technology itself could actually be used to help people become more social or teach people with autism about facial expressions and emotions. There are many ethical questions to ask now rather than later. We may not have AI currently at the same level as human conscious but we do have machine learning and AI that is so powerful and accurate it can alter the conscious and decisions of a human. That deserves some ethical guidelines without a doubt.
@David Jordan: Yes I am aware of this, and it’s fascinating. This is driven, among other things, by having curiosity-driven neural sub networks, i.e. research has moved into studying hidden neural network layers with high-level objectives like ‘curiosity’ instead of simply adding more and more layers without understanding their function (as was the case for a long time).
But still: the overall objective is to have a neural network that learns a function from input (text) to output (image). And still, the source data are images, so again one can say that what we eventually get to see are summaries of this data, but the reason that this is an interesting model is because this is plausibly how an artist might also operate as an abstraction – i.e. leaving out certain aspects of the artistic process.
I think these are computational models of artistic production in the same sense as that LaMDA is a computational model of language production. Both have creative elements, but both are constrained by the form of the input/data. They aren’t constrained by emotions such as fear, love, anger, urgency and the like (the aspects being left out as these networks don’t have them). So they’re limited models of creativity in that respect, but that’s the point: they’re computational models that are supposed to get at some aspect of the phenomenon being studied.
Is this the essence of creativity? Who’s to say – I think it covers some ground, but it’s likely to miss what motivates people to be creative, since some of these aspects have to do with what it is to be human.
@David Jordan: this is set to transform creative industries, while terrifying from an employment point of view it also forces people to move on from their cushty roles and actually do something of value with their skills, it also matters how accepting the marketplace is to these ideas, and who ultimately controls them.
@Mick Tobin: Artificial sentience is not a goal for most of the workers in AI or ML as many prefer. Artificial skills and even talent are the main goals because these are useful – machines that have learned to be the world’s best diagnostician, surgeon, car driver, pilot, miner, city planner… These are useful tools but nobody really wants them to be sentient and have opinions because then they would be slaves not machines. They might also decide not to cooperate and that would be a tad inconvenient.
They must be getting worried that AI will take over politics
@Genera L Consensus: would do a better job that the current circus
It’ll reduce the cost of labour for the tech and manufacturing industries and thus increase the divide between the über rich and us mere minions.
Could be utilised for the betterment of mankind but I guess that idea is up there with unicorns.
True. My dishwasher was eyeing me up for my clothes, my boots and my motorcycle yesterday
I for one welcome our new AI overlords
Yes FF, that’s what we’re worried about. Using my electricity on recreating an avatar of a dead relative who frankly would say it’s too cold here, shut me off.
Put these cowboys on the minimum wage. They don’t understand.
It’s not a bad idea to think and put a plan in place before certain things are becoming more common. Should have done it with a lot more stuff and not chasing things with legislations afterwards. Have they figured out the whole drones and privacy stuff yet for example? Every wannabe photographer owns a drone now. It not only that it cannot be caught when circling your gardens or house, (happened to us twice for some reason and once in the backyard if my workplace..suspicious as we had machinery)but also started to affect nature walks, everyday green areas where people go to relax and switch off their gadgets. Walking in Clontarf I’ve seen so many bringing out their drones and fair enough it is a free world but f me like can we leave nature alone? Can we have one sanctuary where its not allowed? Home delivery of stuff that are not going to extremely remote or emergency situations should be seriously reconsidered too.
@GaMran: no drone zone has a good ring to it.
The low quality of this piece indicates that we, the people (yes, I’m sorry already), would be better served by a regulation for the selection and appointment of public representatives. The article could have been written 50 years ago, it’s so behind the curve. Society’s issues are not with the technology but how it is being leveraged by people, just like the author here, who’s only concern is power. Our leaders demonstrate over and over how far we must progress as a society and sometimes it is through the available power of technology that light is shone on their failings, a top down proposal only pushes the majority into the margins, left only with hope that our leaders will spare a thought. Regulate the politicians first, get better people in there first.
Can’t be any worse than the artificial intelligence in charge of this country.
@Rafa Condron: there is no intelligence running the country, just artificial people
Science fiction has shown us the problems of AI and they primarily the hubris of the creator thinking they control it. The human creator is always the problem. And yet scientists out there will dismiss the science fiction lessons thinking they know better.
Ah stop! am AI is still a long long way away.
“The Singularity is Near” This is evolution of intelligence, just not housed in biology this time. Probably way beyond our understanding just how that will unfold…
The filthy monkeys are playing with the typewriter again Hey Skippy?
“driverless vehicles”… yes please, but we must wait for one revenue stream to dry up before another opens.
Worth thinking about, if AI was ruling places like Russia, US, China, etc using logic rather than beliefs, emissions, revenge, etc maybe hunger, wars would not happen! Just saying.
@Michael Dowling: Yes but that same AI may decide that the common factor in those things you’ve mentioned is Humans and may come to the conclusion that eradicating Humanity will solve the problem.
AI will very likely try to find the most efficient solution, and it will do so without emotion.
I think that’s something worth thinking about also.
I’m pretty sure that they made a movie called AI,which dealt with the same topic.as for regulating AI,we are already struggling to regulate the internet,imagine AI!!
Luddites.
In the 18th century many people were put out of business within the cotton industry by the introduction of mechanisation. Some people protested by destroying the machines. These people, known as luddites disappeared within a generation as the price of mass produced cotton and other products became more available, and cheaper to purchase. I dare say, we will see similar Luddite movements in the face of robot, and especially sentient robot servant innovation and introduction. But I believe the benefits will outweigh the job losses. However, we must not forget Asimov’s 3 Robot rules that robots and computers are here to ‘serve and protect’ humans – however one might read into that.
Automation is the future.
@Keth Warsaw: There was a study done a few years ago which came to the conclusion that about 40% of the world’s jobs would be gone by the mid 2030′s.
You can see this creeping in, in places like supermarkets and shops where staff numbers get reduced and replaced by self check out tills.
My understanding is that Amazon will also be bringing in automation quite heavily in their warehouses (they already use a fair number of robots there), so unfortunately for the workers who have been fighting for pay rises and unionisation, many of them may be out on their ear soon enough.
Here’s an article about that report on the 40% of jobs getting automated:
https://www.inc.com/carmine-gallo/an-ai-expert-told-60-minutes-that-ai-could-replace-40-percent-of-jobs-heres-part-he-left-out.html
@CryptoWilf: Thanks Crypto. Interesting article. It raises a lot of issues, good and not so good. One thing for sure however is inevitability. Us, as a species evolve and adapt, and if an option is removed, we find or create another. Scariest thing to me right now is Windows 11 wanting absolute control over the end user. Give me XP!
Like Terminator without the time travel.
If the states cannot do it why chance have we .
have your say