A Thought Experiment: The Future of AI

AI is attracting a huge amount of attention, and AI companies are attracting a huge amount of funding and very high stock market valuations. So in this newsletter we’re going to examine the future of AI using a scenario planning thought experiment, which will help us consider some of the key possibilities for the future.

We’ll explore, in particular, what would have to happen for it to wreck the economy and therefore society, or boost them to new levels of productivity and joy. 

 For the purposes of this thought experiment, we take it as a given that the technology actually works and will continue to get better. It’s already pervasive, and many companies are investing heavily to try to figure out if they can use it to increase productivity and profitability.

It’s continuing development and recent success is in line with Ray Kurzweil’s thinking as explained in his two books, The Singularity is Near, and just recently, The Singularity is Nearer. Kurzweil is a technologist who argues persuasively that the development of pathway for AI leads to a point at which machines have become “smarter” than people, and all the confusing power and decision-making dynamics that this will entail. “Smarter” is in quotes, because our traditional definitions of smartness will have to be revised in light of what AI and machine intelligence are and perhaps are not. 

 Kurzweil suggests that the smartness threshold (however defined), will be achieved around 2045, at which point we enter a new and entirely unpredictable state of affairs. Hence his choice of the term “singularity,” which in physics describes the unknown conditions in a black hole, where the laws of gravity as we understand them do not apply. His point, of course, is that when machines have become smarter than people, the structure of society will reach a new and entirely unpredictable state of affairs as well.

 2045 may thus be an end point of sorts, and also a starting point, but we’re a long way away from that as of today in 2024.

 For today’s AI industry and for its enthusiastic investors, the pressing question is the business model. That is, how do they provide value, and how do they make money? Right now the firms making the most money from AI are the ones that make the infrastructure, such as Nvidia, which makes the computer chips that run AI. Server makers and cloud platform companies are also doing well, as AI compute requires a big step-up in infrastructure capacity.

 But the companies that make AI tools are not yet making money, and no one knows when or if they will. Hence there is a major business model question – as in, what is the business model? This gives us the vertical axis on our scenario matrix. At one end we see a viable business model; at the other end we don’t. (If you want to learn more about business models and business model innovation, please check out our highly praised book Business Model Warfare.) 

 Society is staring at an even larger and more pressing issue, the potential impact of AI on jobs and the economy. Specifically, will AI become a source of rich and rewarding employment, or conversely a job-killer of epic proportions. As of now, no one knows. So our horizontal axis is the impact of AI on employment, a job-booster or a job-killer.

 These two highly uncertain but rather important questions define a compelling 2x2 grid and that gives us, in the normal way of scenario planning, four provocative possible future worlds to explore.

In world #1 we see a situation in which AI augments human creativity and enhances employment, leading to better human outcomes, and there’s a viable business model, so investors do well also. Everyone wins, which then enables the massive scale and scope of investment that could indeed set the stage for the technological advances that would usher in the singularity. 

In other words, AI is a boon for employment and for investors, until its development brings us to the point of the monumental unknown that the singularity inevitably entails. Does the party continue thereafter? This we cannot know …

In world #2, the employment, productivity, and economic benefits of AI manifest, but a viable business model for AI investment never emerges. In this scenario, AI becomes a philanthropic and government priority, but it never blossoms into a large-scale private sector investment strategy. Which implies less investment in technology, perhaps delaying the onset of the singularity, or averting it entirely.

 In world #3, the AI industry never does find a workable business model, meaning that the industry of AI never lives up to the hype, and it remains only an infrastructure component, but one that that never dominates the economy. Jobs continue to be lost to robots but the whole scale of the endeavor remains modest, and mass-scale adoption never happens. Investment funds dry up and AI firms become just another niche player rather than the tech sector’s headliners. The singularity is similarly delayed or averted, as in world #3.

And in world #4, where AI finds a productive business model, entrepreneurs and investors continue to pile in because it’s clear how to make money. The impact on the economy grows in a productive way. But this is achieved largely by wiping out jobs, a replacement scenario, this causes progressively worse conditions to society, as mass unemployment becomes unstoppable. The technology titans do incredibly well, and everyone else does incredibly badly. Eventually governments must step in to control the damage, introducing who-knows-what type of regulations, and given the vast unknowns, with unpredictable consequences. 

•••

There are a few standout lessons from these scenarios …

 First, the business model really does appear to be a fundamental fork in the road. If the sector figures out a profitable, mass-market approach, then its impact will balloon. If not, the massive market valuations of some key companies, like Nvidia, will follow the well-known boom-bust pattern.

 It’s also worth noting that at this stage of its development, there’s no reason necessarily that AI should have found its best business model. It’s common for impactful and transformative technologies to lack a viable business model for quite some time, even as investors continue to back companies in the space in the hopes that a business model will emerge.

 Google is a good example. It took years for the company to figure out its business model – advertising – and until then it was soaking up venture money and building up its technology and user base, but not actually making money. Once it adopted the advertising model, of course, then it exploded in market value. But that was not an obvious move at the outset (and Google’s founders had even sworn not get into the advertising business)

 Many other firms have similar stories to tell – Fedex lost mountains of money before becoming profitable, as did Amazon.com, and plenty of others. So the lack of a viable business model is not a permanent disability, but it’s will have to be solved sooner or later or the funding will mostly dry up.

And companies like Nvidia, which have ridden the crest of the AI wave to astronomical heights of stock value, will settle back down to a valuation more in line with the past (and a lot of investors will get burned in the process).

 If we make the assumption that AI is already providing value as a part of the infrastructure for a large and growing list of companies, which is what does seem to be the case, then the question is whether it will continue to be an infrastructure element only, or if an AI company can get out front and actually become a mass market thing in and of itself. As of now, no one knows.

AI’s job-killing potential, likewise, will attract persistent attention, and the statisticians will be trying to assess AI’s impact on job losses, job gains, productivity changes, and corporate profitability in order to figure out what real impact AI is having. They’ll even take a stab at assessing AI’s impact on factors like creativity, innovativeness, and job satisfaction, surely a more difficult set of outcomes to measure, but ones that society will nevertheless consider important.

All in all it’s surely an interesting and even important story to explore, and it’s one that we’ll be paying very close attention to in the coming months and years.

 What about you? Do you have experiences of AI that lead you to think that any of the four scenarios described above are more plausible or more likely than the others? If so, we’d love to hear from you! (LMorris@innovationlabs.com)

We called the previous edition of our entire newsletter “Early Warning,” but some readers commented to us that it seemed like a rather ominous and not very optimistic name. We thought that was a legitimate point, and we appreciated the feedback. So now we’re trying “FutureWatch” as the name of the newsletter. 

What do you think? Please let us know …

And despite the name change, we still do need early warnings of impending change, so here are few recent bits of interesting news that have much relevance to our futurist point of view, as reported in The Economist.

Changing: Warfare 

Ukraine has created a new military branch, the first of its kind in the world. Its new “Unmanned Systems Forces” are dealing in drones and electronic warfare, both of which are having massive impact on the conflict. Soldiers and leaders worldwide are paying close attention, as the battlefields of Ukraine are now revealing how new measures and counter-measures define a literal arms race that is also defining the conflict, and foretelling the future shape of warfare. (“Robocommander,” July 27, 2024)

The Chief of Great Britain’s General Staff, General Sir Roly Walker, is working to double to size of Britain’s fighting force within the next three years, and continue increasing it thereafter. Looking forward, he foresees that Russia, China, and Iran may all be in position to start a new round of global warfare as soon as 2027. Following the experience of Ukraine (above), he is also investing heavily in the new forms electronic warfare in order to increase the productivity of his fighting forces. Using technology, he says, “We will sense twice as far, decide in half the time, deliver effects over double the distance with half as many munitions.” Since all nations are playing the same game now, the prospects for ever more deadly and destructive warfare looms ominously. In this case we literally do have an “early warning.” (“Fighting talk,” July 27, 2024)

 Changing: The Energy Mix 

90% of the new electrical general capacity installed in the US in 2024 will be carbon-free. And over the next two decades, America is on track to invest more than $7 trillion in low-carbon energy. So no matter who wins the next US election, the energy transition will continue due to market forces as much as because of regulatory ones. (“A fossil-fuel fantasy,” July 27, 2024)

Changing: Urban Transportation 

China’s Baidu is now introducing its sixth generation autonomous vehicle, and the industry in China is now far ahead of the rest of the world, both in terms of hardware and software. But about 10 million Chinese make their living driving taxis and Ubers, and putting them all out of work is not going to appeal to the Chinese government. An interesting decision point is therefore approaching, one that, as we saw in our AI thought experiment above, may sooner or later bring to many industries in nearly every nation. Is it really going to be a fork in the road, or will we find some way to bring productivity benefits of tech without massive job destruction? That’s a critical business and economic question that every nation may soon be confronting. (“Bump start,” July 27, 2024)

Changing: Elections 

For the first time ever, elections are being held this year in countries with more than half the world’s population, bringing the conflicting impulses of would-be autocrats, democrats, and populists into very visible competition for public support. The French elections in June were particularly notable because the center and left were forced into an awkward marriage to stop the momentum of the right. Quite surprisingly, more than 200 candidates withdrew from their local races in order to concentrate the vote on the left and keep the far right from gaining power. This was an astute political move that undoubtedly required a lot of not-so-gentle persuading, and a clear appreciation of the bigger picture to pull off. Is it also an indication of what will be happening elsewhere as the center and left are confronted with ever more potent populists from the right? (“The elusive art of compromise,” July 13, 2024)

Will Peace Come In Israel and Palestine?

Is it possible that there will be peace in Israel? Opinion polls reveal a disturbing pattern which partially explains why there hasn’t been an agreement. It seems that the approval rating of Hamas increases during wars, and declines during peacetime. Netanyahu’s coalition, likewise, benefits politically from fear and violence. 

It is under these conditions that peace negotiators are working to bring the immediate conflict to resolution and end the stalemate, but as the leaders on both sides know that their support increases during periods of violence, the political incentives to continue the fighting clearly conflict with the social benefits of peace. Meanwhile the suffering continues, and it’s clear that strong leadership and a shift in public opinion will be required to shift course. (“Dance of death,” July 13, 2024) 

In a typical weekly issue of The Economist we find many articles bringing forth useful details key aspects of our changing society, and hence it remains an incredibly valuable resource for its breadth of coverage and its candor. It’s essential reading for strategists and futurists worldwide.

 But useful as it is, it does not convey the whole picture. It’s one thing, and important one, to report on what’s happening, but quite a different matter to put these happenings in a useful context of the broader patterns of societal change. For that you need a framework, or a model, that describes the world as it is now and as it is becoming. And this is exactly what we aspired to do in our most recent book, Hello, Future! The World in 2035.

With detailed discussions around eight major topics that are shaping our future, it’s a detailed study of our changing world. Please get your copy today, available in print and ebook versions at Amazon.

SOME NEWS FROM INNOVATIONLABS

Hello, Future! has received a very enthusiastic response from the initial group of readers, and we’re very encouraged that the book will be successful and even possibly influential.  

Here are a few of the nice comments we’ve received: 

France: Eric Stone
“Compelling, thought-provoking, and brilliantly avant-garde! This book abounds with profound insights into the inner workings of our cultures in their pivotal dimensions about our future.”

Canada: Cyndi Seifried, The Co-operators
”Whether you are a futurist, innovator, systems thinker, or you just care about the future of the world for your children and your children’s children this is a book for you.

USA: Eric Bakey, Boeing 
“Accurate snapshot, balanced research, and a roadmap to create the future we want. Langdon is a futurist, optimist, and a fantastic author.

Spain: Stefan Courser, Professor, Hult International Business School
“Great book for senior leaders navigating our transformative times. Great introduction to the future we are currently facing.” 

To build on this momentum, we're actively looking for opportunities to give keynote speeches and executive briefings based on what the book reveals. We have also developed a new web site – LangdonMorris.com – along with a new set of promotional materials.

Please check them out and share your feedback. 

And if you think this set of topics would be useful or valuable for you and your team, please contact us to discuss a possible engagement.

Langdon will be delivering speeches and workshops in Colorado, Arizona, New York, Ontario, Canada, and South Africa in the next few months, and would be happy to add your organization to the agenda.

 Also, we want to let you know that in addition to the Innovation Mastery online course, we now have a YouTube channel which makes many of the class videos available for free. Check it out @InnovationMastery and let us know what you think.

As always, thank you for your interest in our work.

We love to get your feedback, too.

Contact us at LMorris@innovationlabs.com

If you are interested in inviting Langdon to speak at your organization about preparing for the future, specifically with AI, contact lmorris@innovationlabs.com