#38: Hot Takes on AI (Part II)

More thoughts and predictions

Wow, last Friday’s Hot Takes on AI was a hit. From emails to texts, I received the most feedback than any other piece to date.

Guess what? We’re back for round two, with a focus on AI’s impact on work and the labor market. Keep the replies coming. Heck, drop a comment or a like on the article if you are a fan (or not a fan) of what you’re reading. Feedback helps me get better, which in turn will lead to more entertaining reads for you.

Here we go.

Heat Level: 🔥

AI agent builders will boost product manager productivity by helping them communicate product design and functionality more clearly to engineering teams.

Note: if you aren’t familiar with AI agent builders, check out Hot Takes on AI!

My favorite part of AI agent builders is how quickly I can go from idea to visual concept, especially as someone who’s not a developer. If I have an idea for a software product, I’m able to open up Replit or Lovable and create a prototype in hours, if not minutes. Which is especially useful since my coding skills are rudimentary at best (SQL and a bit of Python).

To be transparent, I’m not actually coding at all when I’m in Replit. I’m typing out instructions in prompt form (read “vibe coding”) and iterating back and forth with the agent builder chat interface on which features to tweak. I even use ChatGPT to write prompts specifically for Replit.

AI helping AI. Isn’t that something, huh?

I believe AI agent builders can rapidly increase the speed-to-market of software products, as they allow those who are not technical to show engineers a clear mock-up of what they are looking to build. I’ve seen non-technical folks build MVPs that help form the thesis of a tech start-up.

Think about it: if a scrappy product manager can vibe code a visual that represents the business requirements of a project, there’s less room for misinterpretation of the ask. This is wireframing on steroids. It will increase the efficiency of both product and engineering teams and speed up the go-to-market process.

Heat Level: 🔥🔥

AI’s biggest workforce impact will come when Corporate America embeds agentic workflows, which I predict will be in 2-3 years.

It’s one thing to say your employees use ChatGPT. It’s another thing to rely on an AI customer service agent. And it’s a whole different ball game when AI is at the core of your operations. I’m talking about true automation of sales, marketing, supply chain, and engineering functions. Where you have fewer, but more technically savvy employees, deliberating strategy and making decisions while the AI does most of the manual work.

I used to work in Corporate America and have experienced how slow-moving big companies are. Even if a large conglomerate is gung-ho about AI, it will take years until that company can fully adapt to changing technological needs. They spend months if not years setting up a “committee” who then works with business unit leads to create a strategic roadmap and execute upon it. But wait, don’t forget the consultants. They’ll be brought in to advise, which they’re doing at a blistering pace. Remember “digital transformation”? I first heard it in college (mid-2010s), and it only faded around 2023. Which is right when AI came into vogue.

I see a barbell approach happening as it relates to AI adoption and subsequent hiring behavior in the labor market. Today, flashy tech start-ups are boasting about their revenue per employee statistics, some reaching into the millions. Not only is technology their core product or service, but they are also building their company from scratch which enables founders to choose an AI-first foundation. It’s simpler to start with agentic workflows, versus having to tear apart an existing process.

On the other hand, you have Big Tech increasing revenue at staggering rates (thinking of Meta in particular) while laying people off in the process. Meta can justify a leaner workforce given the productivity gains they’ve seen from AI across operations, and subsequently, their financial results. The companies that are the most AI-advanced (also are the ones creating LLMs) are best positioned to implement AI at scale. They’ve been researching AI for years, so it makes sense they can be early adopters too.

But that leaves everyone else in the middle. Through my conversations with peers in the CPG and tech industries, AI has slowed hiring plans as companies are more thoughtful of which processes can be automated before adding headcount to the team.

The next step is true workflow automation. Based on the rapid growth of consultants’ GenAI revenue bookings, my guess is the Corporate America committees have been formed. So now, the consultants are now hard at work. But since big companies move slowly, give it another two-ish years until meaningful layoffs become the norm.

Heat Level: 🔥🔥🔥

AI will fundamentally disrupt the labor market, resulting in lower employment needs.

With this level of technological efficiency comes some grim side effects. You see it now, with new companies bragging about their high revenue per employee numbers. That statistic goes down as you hire more people (which is necessary to scale a business), but what that tells us is that companies being born today are early adopters of automated solutions. I believe that this mindset isn’t likely to disappear as the company scales.

Companies fade out over time. Very few brands are around for decades. They go bankrupt, get acquired, or fade into obscurity with stagnant growth. As the new vintage of companies prioritize a new tech-first mindset that only prioritizes hiring when truly necessary, there will be fewer jobs created compared to today’s labor market.

Big companies will also join the party. As they acquire the challenger brands, they’ll justify acquisition premiums by scrutinizing even more about how they can trim expenses using AI, meaning even fewer people stick around post-acquisition.

In most cases, an exogenous shock to the labor market eventually results in a return to baseline employment levels. But given how powerful AI is proving to be, this shock is sparking a cultural shift to how companies approach staffing.

Furthermore, I do expect new industries and job opportunities to pop up because of the shifting labor market. We’ve talked about it before in Technology is Profound, but generative engine optimization (GEO) is the SEO for LLMs and has already led to the creation of hundreds if not thousands of jobs. New industries will keep coming up thanks to AI, but will they create enough jobs to make up for the more labor-efficient broader market? Probably not, or at least not on the same scale.

I will always bet on humans to innovate themselves out of tricky situations. But it’s going to be tough to expect new companies to have the same trigger-happy hiring plans as the incumbents.

Plenty of macro factors will shape the broader labor market too. But I’ll leave that to another post.