If there were two defining themes of 2023, they were (1) Rise of Artificial Intelligence (AI), and (2) Layoffs in the Tech sector.

One did not cause the other, but the contagion of each was so tightly interwoven, that it left an impression. I also happened to be touched by both – I was laid off by my employer last spring, started using ChatGPT in my downtime – was startled by the capability of that tool, and took a course on AI last summer. Now that we’re well into 2024, I’ve had a chance to grapple with these strands, and here is my assessment of how these may ripple through the future.

A WATERSHED MOMENT FOR AI

It was through the meteoric growth of ChatGPT that “AI” became tactile, acceptable, and even fashionable; but the technology that constitutes AI covers a much wider swath than what ChatGPT leverages; and has been around for decades. It has just been quietly nurtured in labs, academic centers, and niche startups. But there was a watershed moment in 2016, which went largely unnoticed. It had demonstrated how this technology had the potential to upend the world as we know it.

In March 2016 AlphaGo, a program created by DeepMind (a startup acquired by Google) was set to play Lee Sedol, the reigning world champion of the ancient Chinese board game called ‘Go’. It was an ancient game that until then was considered by experts to be orders of magnitude more difficult for a machine to play than chess. It was played on a much bigger board, and the combination of choices were magnitudes more than those in chess. Computation needs for a machine to play and beat a human in Go were considered prohibitive – AI models that rely solely on the brute strength for logic and computation were assumed to have a negligible chance of defeating a human expert.

Humans are gifted with this thing called “intuition” – the ability to draw on a reservoir of experience in ways that are not straightforward. Intuition was a considered a critical asset in a game as complicated as this. And the human that AlphaGo was facing was not just an ordinary expert, it was a 9-dan player (the highest level a professional Go player could rise to), and the reigning Go champion.

A set of 5 games were planned, and the matches were keenly followed by AI enthusiasts and by fans of the ancient Chinese game. The odds were heavily in the favor or the human. When interviewed before the competition, Lee Sedol stated that he was expecting to sweep the series 5-0; and maybe if something out of the ordinary was to occur that caused him to slip, then the machine may eke out a draw, instead of losing all five games.

The series started with Lee Sedol playing a move and then looking up to see his opponent, as all Go players who play other humans are accustomed to doing. But the person sitting across him was just an employee of AlphaGo, awaiting instructions for the next move from the “machine”. There was a long awkward pause as the machine calculated and the AlphaGo employee waited with servitude. The pause was so long that he started to look concerned – could the program created by his employer have conked out right at the beginning? Was the game of Go so complicated that the AI behind the algorithm is having trouble narrowing down a feasible next step? But then all of a sudden the machine whirred into action and spat out instructions. The game commenced and the next many moves went along much more smoothly. The little black and white pebbles started to crowd the board. The human champion played with practices swiftness for a while, but at some point there were a succession of moves that clearly bothered Lee Sedol. His demeanor started to change gradually from smooth confidence to focused concern. What happened next was brutal and painful. Sedol’s demeanor started to change and one could see him staring at the board with intense concentration for long durations between steps. And it was a matter of time when the champion realized that he had been bested. He lost the very first game in the series. He was down 0-1.

The next game was crucial. He had to win in order to level the field so that he could then take the set back later. But the game progressed similar to the first one. The score at the end was 0-2. He was in deep, deep trouble. Now he had to come back and win the next three games in order to turn the series around. He had to do it for humanity. The weight of the responsibility was apparent at the press interview after that game. The next day was everything.

The next game was another loss. 0-3. It was certainly a one sided competition, but not in a way that Lee Sedol had imagined. He continued playing the remaining games, but the atmosphere in the arena and across the world wherever viewing parties were being hosted had turned from festive to something like a funeral. In the end Lee Sedol did end up winning one game in the series, with the final score being 1-4 in favor of the machine. That spring morning in 2016 was the day the fallacy was blown to smithereens – the notion that humans had an inherent advantage over machines in intuitive fields was vaporized.

It just took a few more years for that notion to go mainstream. What was demonstrated to that niche group in the Spring of 2016, was made clear to the masses in 2023 with the rapid ascent of ChatGPT. AI had arrived. And it was here to stay (and ready to eat everybody’s lunch).

THE TECH LAYOFFS OF 2023

Moving on to the second strand – the tech layoffs. They were not really related to the advent of AI. Various events of the last few years brewed a perfect storm – the war in Ukraine, the embargo on Russian goods, strained supply chains, increase in gas prices, pricing pressure on consumer goods, lower consumer demand, etc. all causing a gradual upward crawl of the inflation millepede. This in turn resulted in central banks raising interest rates to pump the brakes on inflation. Which then resulted in higher borrowing costs for consumers as well as businesses. The higher borrowing costs on the consumer side resulted in consumers tightening their belts and spending less on products and services.

So corporations were hit two ways – reduced demand for their products/services, and higher borrowing costs for business loans used to grow their business as well as for lines of credit used for financing cash flow. This double whammy hit Tech companies even harder because they had expanded rapidly in the past few years when borrowing costs were rosy. But now they were feeling pangs on both sides of the equation – the revenue and expense sides. And many executives in these corporations resorted to the first tool they could find in their trusted toolbelts – Layoffs!

I will not focus on the number of layoffs, or delve how they effected laid off employees, but will focus this post on the impact on the companies that are conducting the layoffs.

There are a few different myths around how layoffs serve the companies that deploy this tactic:

Myth #1 – Layoffs boost stock price: False. According to Professor Jeffrey Pfeffer of Stanford’s Graduate School of Business, an analysis of 1,445 downsizing announcements between 1990-1998 showed a negative stock-market returns, with more negative effects associated with larger downsizings. Myth #2 – Layoffs improve productivity: False. A study of ~140,000 U.S. companies using Census data found that companies with the greatest increases in productivity were just as likely to have added workers as they were to have downsized. Myth #3 – Layoffs increase profitability: False. Even after statistically controlling for prior profitability, a study of 122 companies found that downsizing reduced subsequent profitability.Myth #4 – Layoffs reduce costs: False. Layoffs may seem to reduce costs in the medium term (after the cost of conducting the restructuring has been factored in). But according to Professor Sandra Sucher or Harvard Business School, companies often overlook the hidden costs. She says that it can take years for companies to bounce back from setbacks like loss of institutional knowledge, weakened employee engagement, higher turnover, and lower innovation.

These are good data points, however I believe they are symptoms of something deeper at play. I believe the root cause that results in this lower performance is related to a commodity that cannot be quantified on any spreadsheet – a concept called Psychological Safety.

Psychological Safety is what allows employees to trust their employer, and feel empowered to take chances (including chances on things that are not sure shots… chances on initiatives that may not seem like they would succeed at first). Even in my book on business lessons from the life of Alexander the Great, I compare the culture in Alexander’s army with that of the Golden State Warriors (the NBA team lovingly known as “The Dubs” in the Bay area of Northern California). There I demonstrate how the key ingredient for the success of both organizations boils down to the trust fostered between the organizations’ leaders and its individual team members.

If culture eats strategy for breakfast, then a winning culture cannot exist without the prevalence of psychological safety across the rank and file. It takes a long time to build that trust – one thread at a time. And that fabric is the first casualty of a layoff announcement. Once employees (who have been fed stories about the workplace being a “family”) see their colleagues set aside at the first sign of economy difficulty, the realization sets in that they are expendable as well. Once that realization sets in, the blanket of psychological safety goes out the window. Everyone feels afraid to stand apart from the crowd. It seems safer to adhere to the norm. Behavior that is considered ‘Average’ seems safe. That would never get them in trouble. And that mindset is the death of innovation.

THE SIGNIFICANCE OF LAYOFFS IN THIS ERA

Let us go back to the match between Lee Sedol and AlphaGo.

Sedol, the Go champion ended up losing the series 1-4 to the machine. He lost the first three games, but continued to play till the bitter end. He eventually won a game (even if it did not matter for the series), and that was the saving grace for humanity. He was not completely outclassed by the machine. The solitary game that he won had seemed like it was also going the same way, until he played a move called the ‘Wedge’. It was the 78th move of that game, and it completely threw off the machine. After that move AlphaGo’s own moves started getting erratic. Many Go enthusiasts labeled that move by Sedol as “God’s touch”.

The reason that move had thrown off the computer was because it had the probability of 1/10,000 for a human to play. What does that mean? The computer was playing by calculating each and every move by analyzing 30 million moves from 160,000 prior games and optimizing for the next move that would increase its probability to win. It was looking at patterns from that massive dataset of prior games and making its decisions. The reason Sedol was able to throw the machine off was because he played a move that was so out of the norm of prior games, that the AI did not know how to interpret it.  

So what does that tell us about the future? What bearing does that have on the companies that are laying off its employees?

Don’t read on.. Think about this for a few minutes. Step away from the screen. Take a walk. Come back to this post later. In 5 minutes, or an hour, or in a few days. Consider how the solitary game that Sedol won against AlphaGo matters to companies that are considering laying off employees.

Yeah?

In my opinion, companies that lay off employees for short term gains risk losing the one edge humans have over machines. By stripping away psychological safety, they push their employees away from taking chances, and towards conformance. They move the organization from being an empowered entity where employees take chances based on what they feel is right, towards a command-and-control structure where they wait for orders from senior leadership.

And with the advent of AI where machines are getting increasingly capable of efficiently harvesting prior knowledge, conformance is better served by machines. What would drive innovation for the next wave will be the “Lee Sedols” in their organizations who would make the 1/10,000 move based on their gut instinct. Maybe human experts working in combination with AI, but we do need those experts to drive innovation – the humans who have divergent thinking and are not afraid to practice it.

IN CONCLUSION

The late American poet, E.E.Cummings once visited a zoo, and upon seeing the polar exhibit there he observed that penguins possess a double existence. He noted that they possess a terrestrial self that is “awkward, ludicrous, ungainly”. In that self the Penguin “cannot fly, and instead walks about, imitating humanity in general and Charles Spencer Chaplain in particular”.

But then he contrasted that avatar of the Penguin with an alter ego of that animal. He noticed how it changed instantly as it dove and sliced under the surface of water. Cummings was startled at that transformation and remarked:

“For whereas terrestrially the Penguin is angular, restricted and sudden, aquatically he is comparably fluent, completely uninhibited, and (when he makes a dart downward through the water in pursuit of his prey) irrevocable. This astonishing self flies through the water by virtue of those very wings most people consider so pathetic and inadequate.”

To me this analogy perfectly encapsulates the hidden costs of layoffs. Layoffs make the remaining employees feel unsafe, and they nudge them to conform to senior management’s expectations of them. They are pushed into remaining confined to their “terrestrial” selves. And the irony is that in today’s day and age – with the advent of AI, innovation requires them to embrace their “aquatic” alter egos.

I don’t envy corporate execs charged with navigating the crossroads of declining demand and increasing borrowing costs. They are given a difficult choice, and laying off employees seems like an easy choice (“easy” not from a moral perspective, but from an economic lens because laying employees off seems like the quickest way to restore equilibrium between market reality and a corporation’s capacity).

But if you happen to be the said exec, faced with this dilemma, please consider the true cost of layoffs for the long term well-being of your organization. Consider not just the 5%/10%/15%/etc. labor costs that you may apparently be able to reduce, but also the loss of innovation within the remaining 95%/90%/85%/etc. of the workforce that is left behind.

There are other options available to layoffs – they may not be easy and they may require more nuance to orchestrate. But I hope you can gather the courage and make the 1/10,000 move that allows your organization innovating well into the future.  Godspeed!

Notes:

Ongoing tally of layoffs: LinkStudy on long term effects of layoffs on mortality: LinkSources of images: Unsplash and Reddit