How to prepare for a jobless Future: Misconceptions about AI’s Impact on the Future of Work
AI as an automated conveyor belt into an uncertain future of work

How to prepare for a jobless Future: Misconceptions about AI’s Impact on the Future of Work

TL:DR // Key Insights (see CoreCortex.ai for audio summary):?

  • ??????????????????????????????????????

To thrive in a rapidly changing job market reshaped by AI, individuals and organisations must prioritise adaptable skills and deep domain expertise, fostering a culture of continuous learning and innovation that leverages AI as a partner rather than a threat.?

Embracing this shift means moving beyond mere survival, to actively shaping a future where human and artificial intelligence collaborate, driving both personal and professional growth in an era of perpetual change.

  • ??????♂???♀?????????????????????????????????????????


Employees and organisations alike are currently facing the challenge on how to get up to speed with the new way of working, alongside our artificially “intelligent” co-workers. And many still feel like they were asked to solve a complex maths equation with several unknowns - without any hint or help where to even begin.

But why should you be interested in AI upskilling in the first place? Well yes, tech companies have started another wave of AI-triggered job cuts, and presumably, this is just the beginning - although some research suggests that “we should expect the effects of AI automation to be smaller than the existing job automation/destruction effects already seen in the economy”.

Job cuts as reported by FT

The flipside of the already observable productivity gains of AI - as Scott Galloway aptly put it - is however a corporate slimming diet that is using the transformative potential of AI for all too obvious business performance enhancement purposes. Based on the simple notion that shareholders will always love an increased output with reduced input (headcount)... ?

Changes in Headcount vs Revenue for Businesses in 2023 as reported by S. Galloway

But where would you actually start after you realised that it is time for you and your organisation to finally do some AI upskilling? Of course there are all kinds of self-appointed “AI experts” waiting to offer their help. Or you could start by reading some inspirational articles, which? unfortunately provide generally less than unhelpful tips, including such clever advice as “Understand the AI Theory”, “Master the Data”, “Learn about AI Tools”, “Enrol in AI Courses”, “Follow AI blogs”, “Read AI Research Papers”, or maybe even “Apply for an AI internship” so at some point in the future you might be able to score a real “AI job”... And of course there is no guarantee that any of the above will actually prepare you for the radically changing work environment trusted expert bodies like IMF and WEF are clearly anticipating.

So effectively lacking the ability to magically foretell how an “AI Future of Work” will actually look like, how can anyone properly prepare themselves for it? One hint might be hiding in plain sight and comes with the insight that, like with many other technological revolutions before, there is no point in trying to foresee exactly what type of new “repeatable, skill-based professional activities” (aka: jobs) will ultimately emerge from the intricately interwoven “effects of augmentation and automation innovations” causing these significant shifts in labour demand. Rather than trying to anticipate or avoid these transformations altogether, we hence might as well fully “embrace” the underlying change itself. Of course not in any obscure, esoteric way, but simply as a logical consequence and in the sense that, due to the rapid technological acceleration over the past years, the traditional concept of fixed, impermutable “jobs” could actually soon become a thing of the past - at least for the parts of society in advanced economies where employment is generally exhibiting a high degree of exposure and complementarity with AI technology.

Matrix-boy telling us about the future of work, created with Imgflip

What that actually means is of course that, in an hyper-accelerated environment of AI-triggered continuous technological advancement, constant change and innovation is becoming the contextual norm rather than the exception. And this is very much rendering any clearly defined job-roles with repeatable, standardised tasks obsolete, at least for the highly educated, high-skilled and digitally literate “elite” able to complement these AI-triggered advancements from the HI side. Apart from having a significant impact on the level of overall inequality, this also means that change literally will become the job of this new class of AI-empowered knowledge and innovation workers.?

This fits right into the recently proclaimed end of career stability which is establishing itself as a new reality for younger workers, who are changing careers much more often than previous generations. And, much like in the context of an ambient AI future of work, this new generation of professionals is strongly relying on their flexibility and transferable skills when they are pivoting from old and adapting to new careers. In a recent interview, David Mallon, Managing Director @ Deloitte is describing the skills needed to withstand the adaptive career pressure from AI as uniquely “human capabilities”, specifically referring to relationship building, curiosity and creativity. In his opinion, the best way to protect one’s job against the destructive forces of AI is to “figure out how to use AI to reinvent what you do. If you're on the forefront of re-authoring your own job, you've increased the likelihood that you're going to be just fine no matter how this plays out.” Again, this sounds very much like the future of work will be a lot about change - and that work itself (and we with it) will keep changing.

The most striking contextualisation and best explanation of the disappearance of traditional jobs in favour of a much more fluid “change management” role however can be derived from the so-called “Turing Trap”. Referring to the now over 70 year old idea originally articulated by Alan Turing, to create machine intelligence that matches that of humans. Whilst this self-referential way of thinking about Artificial Intelligence has since been incredibly successful in driving technological progress with regard to machines replicating human capabilities, the very same progress of course also sparked a wave of automation-based efficiency optimisation, effectively reducing the need for human labour. The idea of a “Turing Trap” refers to us still being mentally hung up on this self-referential “(GOF)AI” mindset from the past, whilst further advances in AI technology have since uncovered completely new ways of using machines together with humans to achieve tasks previously inconceivable.

Visualisation of the “Turing Trap” by E. Brynjolfsson

In more specific words: Our way to respond to the impending AI work transformation challenge is still mentally caught in the ?Turing Trap“ of trying to defend our continuously shrinking human uniqueness against the ever increasing human-like capabilities of machines, instead of trying to look beyond what we once were only able to do ourselves. But overcoming this biassed mindset of thinking about our future of work ?ex negativo“ also involves embracing completely new tasks. Tasks that will often require radically new ways of thinking as much as they will require accepting the constantly changing nature of work alongside continuously evolving machines.

DeepMind's Professor David Silver describes AlphaGo Zero

When looking for orientation on how to successfully make the leap towards this much more progressive, future way of thinking and working alongside artificially intelligent machines we might find inspiration in reference cases where these machines have managed to discover new knowledge and previously unknown problem solving strategies in a completely new, tabula rasa approach which has been entirely unconstrained by the limits and biases of humans.

Progress of AlphaGo Zero, source: DeepMind

Similarly, inspiration might also be found in use cases where medical researchers are trying to build explainable AI models that are not only accurate, but also help subject matter experts like oncologists understand how respective AI systems actually identified these new diagnosis or treatment opportunities, so that the same experts can later re-integrate these new findings as part of an updated medical standard of practice and teaching. Looking into the not too distant future we might even soon come to a point as human species where we will pass on such newly co-developed knowledge, strategies and ways of thinking as part of a completely new way of human-machine co-evolution. This will mark an important evolutionary inflection point, effectively highlighting that we will have evolved to become an entirely new biological species of intergenerational hybrids - the first new hominin speciation event since we parted ways with our closest evolutionary relative, the Neanderthal, 40,000 years ago.????

This sheer unlimited horizon of future co-evolution and co-development with AI is also why a purely issue-focused AI upskilling approach will most likely fail, namely as it is only focussing on compensating for the current shortcomings of such “Turing-trapped” human capability imitating AI approaches, whilst completely ignoring all the potential new, ingenious ways in which AI might be able to thrust ourselves beyond the limits of our current knowledge, understanding and ways of interacting with the world.

Problem based upskilling, based on the pitfalls to watch out in AI adoption:?

  1. Uncharted territory problem: misusing AI beyond its capabilities will backfire, hurting your performance or customer experience.?
  2. Getting-lazy problem: the better the AI systems get, the lazier we humans tend to become in verifying their output.?
  3. Mediocrity-problem: while AI systems are becoming increasingly creative, their output still tends to be rather generic and bland.

(Adapted from T. Vilkamo’s categorisation approach)

Of course that doesn’t mean that there are no approaches for an upskilling process that will allow us to actually identify better ways to complement and integrate with artificially intelligent systems. One of the economically and historically most informed approaches comes from MIT economics professor David Autor, who is stressing that important innovations like AI have historically “never been about automation”, but rather “opened fundamentally new vistas of human possibility” and simultaneously “generated new employment and new demands for expertise”. The most valuable advice Autor is giving in terms of AI upskilling is that of a strong focus on building actual expertise, specifically by showing us what hangs in the balance:

“AI poses a real risk to labor markets, but not that of a technologically jobless future. The risk is the devaluation of expertise. A future where humans supply only generic, undifferentiated labor is one where no one is an expert because everyone is an expert. In this world, labor is disposable and most wealth would accrue to owners of Artificial Intelligence patents.”?

A second, but in no way less impressive guidance on how to prepare for an AI future of work comes from the much respected CEO of the trillion dollar AI company NVIDIA, Jensen Huang. Huang recently mentioned in a public conversation that while computer science and programming were once must-have skills, the focus should now shift to developing actual expertise in relevant domains such as biology, manufacturing, farming or education, mainly because everyone will become a programmer thanks to AI.

Jensen Huang about future work skills on Twitter

In summary, we can conclude that, not only is the fear mongering about AI threatening the future of work misplaced, but it is also misleading and based on wrong assumptions about the general societal effects of major innovations like AI. Rather than anxiously and mindlessly chasing after short-term hypes, tools and trends around AI, any actually valuable upskilling would rather fulfil the following five core-criteria:

  1. Focussing on expanding one’s existing capabilities and skills (together) with AI, rather than trying to defend any “uniquely human” competencies against continuously advancing AI
  2. Emphasising the development of general and transferable skills over more niche or nuanced skills with the aim to increase overall job flexibility
  3. Embracing a context of (accelerated) change by not clinging too much on fixed roles, job descriptions, titles and hierarchies but rather prioritising actual and pragmatic problem solving
  4. Developing a resilient and adaptive mindset along with the ability to constantly reinvent work and one’s role in it in light of the constantly evolving nature of the underlying AI context
  5. Building up a strong and trusted foundation of practically tested expertise in an area like biology, medicine, manufacturing, farming, education or any other practically relevant domain?

Out of this list of requirements, the very last one, stressing the importance of actual domain expertise, will most likely be the hardest to achieve. Especially as you might be able to “automate automation”, but you simply can’t automate building up domain expertise. Which might warrant it to also provide a few principles specifically aimed at building the strong foundation of domain expertise required for flourishing in an AI future of work. Here is an attempt to do just that:

  1. Learn how to always apply Critical Thinking
  2. Learn from First Principles
  3. Learn from own Experience & Interaction?
  4. Learn by Solving Real Problems
  5. Learn by Failing (fast enough)

Ultimately, and without diminishing the importance of any of the above criteria and principles for AI upskilling, developing a better understanding of what AI technology actually is (and will soon be) able to do might generally be a sensible advice in order to make more informed decisions about how to prepare for a work environment where ubiquitous AI will soon be the norm instead of the exception.

要查看或添加评论,请登录

社区洞察

其他会员也浏览了