Will AI be able to do everything?

This is part two of a look at the BCS ‘Digital in Business Life 2024’ survey, where we asked responders to comment on Rishi Sunak’s chat with Elon Musk, where the self-aggrandising rocketeer said: 'There will come a point where no job is needed — you can have a job if you wanted to have a job for personal satisfaction, but the AI will be able to do everything.'

Cult of personality

As you may expect, we had a lot of comments on this topic – many of them questioning Musk’s credentials. We make no comment on those, however, this was intended as a jumping-off point for a more informed debate.

In an overview of the verbatims we had in answer to the above question, ChatGPT informs us that the member responses: express scepticism about the idea that AI can replace all jobs or that we would want it to. There's a recognition that while AI may automate many tasks, it cannot replace the human elements of creativity, emotion, and personal connection that are important in many roles.

Overall, the consensus seems to be that while AI will significantly impact work and may automate many functions, there will always be a need for human jobs, especially those requiring complex, creative, and interpersonal skills. The transition to a heavily AI-dependent society is seen as complex and fraught with challenges that need careful consideration beyond just the technological capabilities.

The large language model also analysed these comment to produce a list about whether AI has the capacity to do everything. Here are the main points extracted from the comments:

1. Economic implications: There's a belief that unless we completely overhaul our economic system, people will still need jobs to afford essentials like food and housing. AI is not seen as capable of replacing all jobs, especially those requiring personal interaction such as nursing or teaching.
2. Importance of human interaction: Many responses highlight the significance of human interaction in many jobs, which AI cannot replicate. Despite AI's growing capabilities, there will always be a need for physical human presence.
3. Job creation vs. job displacement: Historically, technological advancements that create efficiencies and automate tasks have also led to the creation of new jobs. This pattern is expected to continue, suggesting that AI will change the job landscape rather than eliminate it entirely.
4. Role of AI in decision support: In the near to medium-term, AI is expected to grow more in supporting decision-making roles, with humans remaining as the final decision-makers.
5. Scepticism about AI replacing all jobs: There is scepticism about the idea that AI can replace all jobs, with some calling it scaremongering. Comments point out that politicians and public figures often lack a deep understanding of AI and its limitations.
6. Technology shifting job roles: Historically, technology has shifted jobs from manual to skilled ones, and AI is expected to continue this trend, not necessarily eliminating jobs but changing their nature.
7. Concerns about AI overreach: There's a fear that if AI develops beyond human intelligence, it may lead to a loss of control and potential societal upheaval due to the disparity between those who control AI and those who don't.
8. AI as an augmentation tool: AI is viewed as a tool that could free up human resources to do more interesting work by taking away mundane tasks, rather than completely replacing human jobs.
9. Societal and ethical concerns: The prospect of AI taking over jobs brings up societal and ethical concerns, including the potential loss of freedom, control, and choice on a societal level.
10. AI's limitations and governance: There is a consensus that AI will have limitations and will require governance, including laws and treaties, particularly concerning its use in sensitive areas such as military applications.
11. Elon Musk's comments on AI: Some responses specifically criticise Elon Musk's statements about AI, suggesting he lacks verifiable knowledge about the subject and recommending that caution is exercised against taking his predictions at face value.

These points reflect a mixture of cautious optimism about AI's potential to contribute to society and deep concerns about its ability to handle complex, creative, and emotionally driven roles. There's a clear recognition that while AI will transform many aspects of work, the human element remains irreplaceable in certain domains.

But how were these views actually expressed by people?

Getting personal

Without putting too fine a point on things, a number of words came up in discussion of the space-faring protagonist’s expertise – among these were ‘b*llocks’, ‘w*nker’, ‘bullsh*tter’, ‘pothead’, and ‘loony’.

You don’t always get that type of flavour with large language models. However, let’s take a look at some of the more insightful comments on areas of concern. The following are lightly edited verbatims.

History

‘Throughout history, the introduction of technology has come with predictions like this. They have always proven wrong. What DOES happen is a shift in workforce to new roles.’

‘Elon Musk's only mode of communication is hyperbole and bombast so I'm not sure we should be taking direction from him. There are many examples from history of new technologies which threaten to/promise to put an end to jobs and working, but humans always adapt and new roles/industries arise. AI will be a turning point but I doubt it will play out as Musk claims.’

‘This type of statement is made after each iteration of improvement in IT, all that happens is a whole new set of jobs emerge.’

‘This same comment has been said about every automation technology, but what actually comes to pass is that the automation becomes another tool that human workers use. It's more likely that new jobs will start to exist that didn't before, some jobs will cease to exist or become much less common, and many others will have elements of working with AI but otherwise will be unchanged. Also, Elon Musk is wrong about a lot of stuff. Have you seen what the man posts on Twitter? Wow.’

‘Many years ago, the arrival of personal computing was supposed to free people up and allow more leisure time. The opposite happened. Greed and the constant drive for efficiency wins every time, sadly.’

Specificity

‘It shows such a fundamental lack of understanding of both people's needs and AI's capabilities. I love AI, I work in it but I don't want an AI librarian - I want to enjoy talking and sharing personal, meaningful recommendations. I want my bank teller to truly, honestly feel my pain and want to help not just have a chatbot that sounds like it. And that's even before you address the issues that AI simply cannot and will not ever be able to do, which is most things. Like all tools, AI is fantastic when it's applied to the right job. Also like most tools, if you try to use it in the wrong place it will at best do nothing, and at worst will cause major harm.’

Human roles/skills

‘I think that there will be some jobs that will still be done even though AI may do most work people will be more comfortable with a person in the loop as the last line of defence. There should be a right to appeal to a human in the justice system and also a human backstop pilot in air transportation - especially with passengers on board.’

‘Hydrocarbon accounting is a blend of chemistry, metering, unique physical characteristics of each installation, changing regulation and.... negotiation. Can’t see AI replacing the negotiation.’

‘There will always be roles that cannot be automated and which computers / machines are not so good at, people will be needed for these roles. Many of these jobs will not be down to "personal satisfaction", many will be unpleasant and need significant incentives. In addition, if no job is needed - where is the money to live going to come from? What about the increased crime that will occur from too many idle hands, the lack of life satisfaction and dignity that is required from useful work, or the mental health issues what will develop because of this and due to a reduction in human to human interaction? Such claims about human work being no longer required have been made during every other significant industrial change in the past and did not come to pass, I can't see that this pattern will be different this time around, the result will be a significant change in the types of jobs available, not the total number of jobs.’

‘Engineers keep the world ticking. We strive for innovation and are hindered by under-qualified programme managers. There will come a day where programme managers can be replaced by AI, and I fully welcome this.’

‘I've worked in psychiatry and - with the best will in the world and while acknowledging that there are AI-based therapy intervention applications being developed - I don't think there will ever be a time when people aren't involved in the care of other people. Also, AI-generated software is possible, but it needs human intervention as well.’

‘AI is unable to improvise, think intuitively and make leaps in imagination in the same way as the human mind. It attempts to emulate these abilities but is limited to its specified programming parameters.’

‘AI can't rack a server, AI can't install power cabling, AI is only as good as the people building it, so this statement is inflammatory headline-grabbing nonsense.’

‘You will still need a human in the loop to make the final judgement call that the target is legitimate and can be engaged safely.’

‘As with all technological enhancements, AI will change where the skills are needed. Learning the skills to successfully prompt is essentially a new paradigm of programming and I believe that you are still going to require people with understanding of code to debug / understand where and why things go wrong.’

‘Total rubbish, good luck getting it to ethically farm your meat and fish. Also it's energy hungry so will need more time and should only be used for large scale projects.’

Human nature

‘There will always be a need for human interaction and I think as advanced as machines can be, there will still be a desire for human contact and a human will be needed if the AI goes wrong!’

‘The climate will get us before AI does at that stage. After that there will be a lot of work and it will all be hands-on.’

‘It sounds great but it is not in the nature of man to rest because he is fundamentally configured to do better and have more. There are fundamental human values that are not about intelligence human or artificial. They all centre on a sense of value and community not the recognition of patterns nor the fulfilment of "work". And the clear evidence of the human race today is place and community. Whilst we have countries, parties, white or black, man or woman or anything we consider makes us distinctive then we will all have our "own" AI to promote this sense of place and community. When someone can make the United Nations work for the whole good of all humanity then AI will just be another differentiator - and we know just how the atomic bomb failed to achieve unity!’

Timescales

‘It is too far into the future, no one from can live enough to validate the statement. It is like saying the earth will be some day swallowed by the Sun.’

‘It may be true, but is it true in any timescale relevant to me?’

‘We are incredibly far from that. I am certain that we will be able to automate and reduce the costs of most of the things that we find valuable nowadays, but as far as money exists, jobs will exist. If something is entirely provided by a self-sufficient machine, then, that has no cost. Open source communities and ONG will be able to produce them and everybody will be able to benefit freely. It feels very far from a world where anyone can get anything that they ask without needing another human to achieve it. In fact, it seems that we are going in the opposite direction, and we are increasing the demand from services where a human is present.’

Embodiment

‘It will take a long time before AI is able to perform physical tasks as well as people.’

‘We live in a physical world - to imagine that AI or other tech will, in even the medium-long term, obliterate all jobs, is to be grossly ignorant either of the technology, or of the importance of people who will ultimately do jobs that require some kind of physical effort. Ultimately, Musk is an appalling aspect of a broken society.’

‘The T-1000 is not going to be emptying my bins any day soon.’

‘I'm still waiting for the machine that is able to take my laundry from the basket, clean it, and place it in my closet.  Until I see that, I won't believe a single word from that man.’

‘AI will decimate the office workforce but it will lag behind when it comes to practical trades like building services.’

‘AI can't cut your hair, build a house, entertain an audience (the entertainment is in the performance of the human), etc. However, sadly, many jobs will inevitably be replaced with AI without any intervention.’

Societal and economic challenges

‘It's hyperbole from an epic showman in Elon Musk. I think we are so far away from having general AI, that to make that guarantee now is nonsensical. We should be working towards a future leveraging a high level of AI, a day and age in which we no longer need jobs is not an AI problem, it's a societal problem which requires things like universal basic income. There are huge societal challenges to solve before we get rid of jobs.’

For you

Be part of something bigger, join BCS, The Chartered Institute for IT.

‘In order to achieve this we'd need something akin to a generalised AI. Given we still don't fully understand how the human brain works let alone how that leads to consciousness I can't see us successfully developing a generalised AI except by freakish accident any time in the near future. I think it's a race between the birth of that kind of AI and humanity's demise due to climate change. 50/50 at best.’

‘No job = no money = no economy ... never happen.’

‘Unless we completely overhaul the way we do economics, people will still need jobs in order to pay for their food, housing, etc. Equally, AI will not be able to take the place of nurses, teachers (especially early years), childminders, etc. so there will not be a point where absolutely "no job is needed".’

‘This is a highly disingenuous statement, as it implies that there will be some sort of safety net such as universal basic income, and that all jobs are capable of being automated. There is plenty of evidence of automation putting people out of work and erasing whole categories of employment, whilst offering opportunities for enrichment for a lucky few. In the UK specifically we also have plenty of evidence of indifference and even active hostility on the part of the state towards people who are not able to be sufficiently economically productive for their existence to be considered worthwhile, e.g. due to debilitating chronic conditions such as long Covid. In any discussions about the "transformative potential of AI" we should be careful to mentally reframe the conversation as being fundamentally about automation and its potential benefits and drawbacks, to avoid being distracted by sci-fi visions of the future. It's also crucial to understand where responsibility sits when machine learning techniques are being used, and the nature of any consequences that people and organisations might face. A particularly good example here is the hype around possible future artificial general intelligence versus the reality of machine learning models being used to guide bombings and missile strikes on the civilian population of Gaza. Who will be held responsible if an ML model commits a war crime?’

‘AI can't really care or have emotions or empathy - AI may simulate everything but it doesn't have everything. As people age authenticity is more important. We may very well see a return to the artisan. Every change causes the rise of its counter change - such is human development.’

Nuance

‘Both Rishi and Elon know nothing about AI, and should not be making statements like that. Just even reading that statement is cringe. My team invested in AI/ML startups pre-pandemic, and I cannot imagine a situation where any of the researchers and developers with doctorates and published papers on the topic would say something as incredibly vapid and ill-informed as that.’

‘I spent nine years studying the field of AI from a general perspective and have a good grasp of what it offers. Statements such as this represent a particular world view. What it fails to recognise is the complexity and beauty of the systems in nature including ourselves and that this complexity can’t simply be reduced to combinations of patterns or solutions. The natural conclusion of this position is in fact the loss of freedom, control and choice on a societal level. Not the utopia many think it will become.’

‘AI is just a tool, the threat of disruption is from capitalism and political failure.’

‘Generative AI is an accelerant - not a panacea. It is so pervasive currently because it is so accessible to the general populous; it is "the iPhone moment" for AI. That said, many feared that the creation of the ATM would spell disaster for warm fleshy bipeds working in banks - that did not transpire and, as such, I don't agree with Musk.’

Read the previous article on this topic