Jolly's Volley: The Pending Cliff

Recently, I was on my way to an MIT event in San Francisco on the topic of “Future of Education in the Age of AI” when, while stopped in traffic on the Embarcadero, an ad at a bus stop caught my eye. It simply said “hire no humans” in big letters with the face of a AI-generated female avatar underneath. As a VC in the age of AI (and in the heart of the AI universe – Silicon Valley), I understood clearly the message that the company was trying to convey. But the tag line (my guess is, by design) generated an almost visceral reaction as the topic of AI’s impact on jobs has been front and center for me, especially as a father of three. Interestingly, the topic of conversation at the MIT event turned into a mix of education curriculum in the age of AI, but more importantly recognition of the fact that college kids, MIT included, are feeling uneasy about their job prospects, especially entry-level white-collar jobs. The juxtaposition of the ad with the topic at the event was a “things that make you go hmmm…” moment for me. Interestingly, right after the MIT event, I headed to the Chase Center for a social event being held by one of the top tier VCs where again, not surprisingly, the topic was AI, AI and more AI. But in this case, the place was buzzing with a couple of hundred people including young 20-something college drop-out YC founders intermingling with a few successful serial entrepreneurs, investors, service providers and ecosystem participants. In other words, it was a typical Wednesday in Silicon Valley!

The current geo-politics has created massive tail winds for AI-native companies (especially those at the intersection of AI and Defense/Robotics), but at the same time has also elevated the level of uncertainty, for the young and old alike. The war in Iran caused Iron Pillar, my firm, to postpone our AGM this year. At that annual meeting, the team, especially my partner Anand, provides a compelling, data-filled overview of the “Indian macro story”. The deck covers trends in Indian demographics, commerce, infrastructure, capital markets, domestic consumption, financial markets etc.  For me, one of the most intriguing charts in the deck is one about India’s population shift expected between now and 2100. India is projected to have its population grow from ~1.4B currently to somewhere between 1.7B-1.8B by 2050 before the number starts reducing. That’s an additional 300-400 million people (roughly the population of the US) over the next 25 years. What’s even more remarkable is that the Indian working age population in 2050 will be ~1Billion!!! Let that sink in…The question that I ask myself is…in the age of ubiquitous AI and accelerating automation, what will those billion people do in a country like India. Layer on that the fact that the percentage of the Indian population that is currently sustained through agriculture (~800M) is reducing as more young people move from rural to urban areas in search of better education/jobs, and are less inclined to stick to farming for sustenance. The VC in me loves the AI boom as portfolio companies have become a lot more efficient and the revenue/employee barometer continues to move up and to the right. But the philosopher in me (I guess) is worried about the proverbial “cliff” that society is moving towards just from an employment and subsequent quality of life point of view. In the world of trillion-dollar IPOs, there is insane amount of wealth being generated for the very few, while most of the global population is either clueless or incapable of adapting to the impending AI tsunami. The question, therefore, that I grapple with is “Are we going to be able to provide a decent quality of life at sustainable income levels for vast swath of global society in a post AI era. If so, how?”.

One thing is clear to me. That we will all be a lot more efficient in doing what today are regarded as mundane professional and especially personal chores (planning travel, reservations, commerce, documentation etc.) That extra time can be utilized for potentially doing other things that we enjoy, learning new skills or who knows…more human interaction for a change. It is also clear to me that certain functions are not going to be as impacted by AI. Tasks such as recruiting and sales will stay very human centric, at least for the foreseeable future. Healthcare, I believe, especially in aging parts of the world, will continue to be primarily human led. Compassion, empathy, feelings, are still very human aspects imho. Then there is my handyman, Mr. Hung, or Leo, the plumber, who do not see their business getting disrupted any time soon. And they have increased their hourly rates:)

I do predict that eventually, most people (and ideally all) will have a personal AI avatar, a highly personalized agent who will be our guide that provides real-time analysis of an individual’s strengths, weaknesses, gaps, and generate curricula or dynamic “to-do’s” to address those gaps. We are entering a world of liquid learning, where we will constantly be unlearning and relearning to stay relevant (professionally speaking). Maybe, just maybe, that’s the answer to my paranoia about “what will a billion people in India do in the age of AI”. Perhaps the collective free time will be optimally used to “make the world a better place” and solve real problems as a society. Perhaps there will be a business model attached to those endeavors. Problems of water, energy, climate, cancer, neurological issues, famine, pandemics…the world does not have a dearth of issues. I am not sure we can ever solve the ultimate human issue of greed, power, wealth, ego…some combination of which is and will continue to be responsible for global turmoil. That has been the case in the past, and so shall it be in the future, as long as humans are involved. Btw, there will have to be justification for the hundreds of billions and eventually trillions being invested in defense related technologies. To generate the ROI, unfortunately, conflict is necessary. We, as human beings, are capable of creating conflict where none is necessary (in part, to justify that spend – the so-called global military industrial complex, now turbocharged with all things AI – digital and physical)

I am usually a “glass half full” individual. But I must be honest…I have been struggling a bit recently. The world is a complex place. The west is developed, supposedly rich, but aging. Parts of Asia and Africa are developing, poor, but young. They both have their unique set of challenges. Coming back to the topic of AI, perhaps the aging and old don’t care as much as they are either retired or retiring soon. The young, at least those who are well educated, might be alarmed and dealing with anxiety, given the AI headwind. But they are also the ones who are likely to learn the new technology and tools faster (heck, they must) to become and stay relevant. But it’s people especially in their 40’s and 50’s, in impacted industries, who are likely to be obsolesced without an obvious off-ramp. Many of them don’t fully realize the dislocation that is happening yet. A slight digression…I still remember vividly when I first entered the job market back in the early 90’s. I worked at a defense contractor in the Boston area. Democrats were in power at the time, and it was clear that the defense budget was going to be cut, leading to layoffs at the company. There was a gentleman in his 50’s who was in the finance group. He knew that his role was likely to be impacted. He was old school, typing slowly at his terminal, more comfortable with a paper and pencil than anything electronic. But to try and persuade senior management to spare him, he worked long hours and would even come in over weekends. I could sense the desperation on his face and remember him telling me that he would not be able to find another job if he was let go. He had been with the company for 20 years. But he knew (in his core) he was obsolete at that point. A young person, half his age and salary could do twice the work in half the time. He was not spared and was let go the following week. I am not defending or condemning the company’s decision, but the reality is that the human toll of the impending disruption, especially in the short to mid-term, is likely to be very high, especially for those caught in the middle of being relatively highly paid mid to senior management but inefficient in the world of AI agents where human productivity is going to be quantifiably measured in real-time. And those who are simply not productive or productive enough, will be let go.

There are many, usually within the echo-chamber of the AI ecosystem in places like Silicon Valley (startups, investors, beneficiaries) who usually point to historical technology inflections, and indicate that the world was better-off post inflection with a lot more jobs created through those very technology innovations. While that is true, I would argue that the scale of the AI-led disruption is much larger this time around, at a time when the global population is also higher than the industrial revolution, advent of the computer or the Internet. I could see those shifts being enablers a bit more clearly than I do the current AI wave. Perhaps that’s why hindsight is 20/20, and one is a lot more concerned when one is in the thick of a disruption and able to see clearly post that phase.  

A thought bubble, however…Every week, there seem to be announcement of layoffs by some of the biggest tech companies in the world. Recently, Meta has laid off close to 20,000 of its staff. Microsoft offered 7% of their staff a retirement package. The decisions are being made in the interest of greater AI efficiency (and pressure from financial markets). They are all chasing the Revenue/employee benchmark of $1M (or ideally over $5M as is the case for the AI-native companies likes of Midjourney, Cursor, Anthropic and OpenAI). The only way to achieve that lofty target is by either increasing revenue or reducing headcount. The latter seems to be a faster path for most companies. It might be an interesting experiment and potentially a branding coup for those same companies to create a program to re-train and re-deploy those impacted employees. The quarterly earnings targets and shareholder expectations may make that experimentation impossible, but the optimist in me feels that there should be a way for the likes of Meta, Amazon, Microsoft and others to create a win-win here by upskilling and retaining at least portion of those employees who are otherwise being laid off.

But that same challenge of “what are a billion people going to be doing” is the flipside of an opportunity. It will be entrepreneurs rather than governments who will address that issue head on. As I think of the opportunity set over the next few years, I divide the market into three broad segments:

1.        Enterprise. There are only 2000 Global 2000 companies:) These markets are very large if companies can successfully address this “top of the business pyramid” segment through the AI transition.

2.        Small to mid-size enterprise: This is a very large swath of enterprises from mom-and-pop shops all the way to the Global 2001. This could probably be compartmentalized into additional sub-categories. But for the purposes of my argument, let’s keep this as a single segment.

3.        Consumer: This, of course, represents opportunities to make AI truly ubiquitous at the individual level.

All three categories above provide massive opportunity to help the respective groups transition to an AI-centric world. Clearly, many companies are focusing on the first category, for obvious reasons. There is a lot of money at the top of this “customer” pyramid. I would argue, however, that collectively, there may be more money to be made in the other two layers where the numbers (of customers) are much higher. But GTM will have to be nailed to achieve scale make the economics work. I personally think there could be a large army of young people who could and should be AI consultants for the long tail of the enterprise (SMB) segment. Those salons, spas, restaurants, retailers and barbershops are not going to be able to figure this out on their own. And while folks may dabble with Claude, GPT and Gemini, it’s unlikely that they will be able to create, deploy and manage bespoke Agents for their respective businesses over the long term. A massive army of specialized FDEs might be needed.

And while the consumer segment might be the hardest to crack, I believe that a few very large companies will be created in the realm of personalized agents across the spectrum – from creation to management to social media, personalized commerce and monetization. The thought of a personalized AI avatar is no longer a dream. Everyone will have an agent that tracks, guides and becomes an extension of the individual. Eventually (and likely in the not-so-distant future), an individual will be able to verbally instruct the agent to do almost any task. My hope is that enough advances are made on the energy and energy efficiency front to not destroy the planet in the name of productivity. And then there is a thorny issue of AI governance and security (I will write about that topic separately). Security, however, is a massive game of “whack a mole” where vulnerabilities and gaps constantly shift. Any time there is a transition to a completely different paradigm, like the one we are going through currently, these security issues (and therefore entrepreneurial opportunities) are bound to emerge.

I still worry, however, whether enough well-paying jobs will be created by AI to overcome the hundreds of millions who might be disrupted by it. There will be ebbs and flows, vertically throughout the socio-economic strata, across different age groups and finally, across geographies with the developed and developing markets experiencing their own set of challenges and opportunities.

India, for one, is taking the right steps, in my humble opinion, to prepare for a future with a focus on both controlling its own destiny and enabling the tens of millions who enter the work force every year. The “make in India” initiative with a focus on so-called “deep tech”, manufacturing and innovation led economy is commendable. And the government is putting its capital behind these initiatives with tens of billions earmarked for startups laying the foundation for the country in areas of innovation including space tech, defense, energy transition, battery technologies, robotics, industrial automation, biotech/lifesciences, semiconductors and other applications of AI. While India might have the resources including intellectual talent, aspirational entrepreneurs and capital, many other parts of the world don’t. I am not sure whether the advent of AI pushes those countries further apart from the rest of the world, or AI provides unparalleled technology democratization for young budding startups anywhere in the world to be able to build world class tools, platforms and companies. Only time will tell:)