The Agents Are Coming: Implications Of AI’s Product Cycle On The Future Of Work

Thank you all for subscribing, following, and supporting my newsletter. The community here has grown to over 11,000 members. At events like WEF Davos, I’m reminded of your power. The newsletter opened the door to my first invitation to speak at last year's event and my return this year.

Many of you helped spread my ideas, and your support over the last 3 years means a lot to me. These articles are disruptive, and the community’s strength comes from its willingness to share, synthesize, and add to them. People who began as followers are now influencers in their own right. We’re creating network effects, and the community’s reach is growing.

We have complex discussions that don’t happen anywhere else, and those ideas are taking hold. BCG (and many others) are picking up on them. In its 2025 AI Radar survey, it found that 75% of executives rank AI as a top 3 strategic priority, but only 25% see significant value from AI. We’ve been discussing the ROI disconnect since the beginning, and it’s a frequent topic during office hours.

BCG’s solutions align with my frameworks and contradict conventional, often self-serving consultant advice. It recommends against diluting efforts and investments across endless pilots and PoCs. AI’s NFT phase is coming to an early end as more consulting companies follow an outcomes-based approach.

BCG found AI leaders who successfully realized value focus on fewer use cases and develop products that go deeper into each one. My frameworks’ focus on vertical depth over horizontal breadth is a key advantage. Most of us inherit a mess of technical, strategic, and cultural debt or are brought to clean it up. Frameworks are why we can do it while others fail.

As more business leaders follow our community’s advice, ROI will rise, and greater investment will be justified. That’s where the horizontal breadth of an AI platform becomes feasible. AI’s success creates a virtuous cycle, and businesses are looking for people who know how to start the flywheel.

It seems like every day, another person rebrands themselves as an AI strategy expert on social media and launches an advisory service. Davos was filled with newly minted AI strategists eagerly skating on the ice between venues. Their optimism is warranted, and the future is bright for people who can turn technology into value. If you had to choose a bandwagon to jump on, value-centric AI and the outcomes economy are the best performers.

It’s a fantastic validation of the hard work people in this community have been doing. The hype cycle is finally leaning into value creation. AI strategy will lead to more products and platforms hitting the market. The future of AI is increasingly pragmatic, but that doesn’t mean it’s bright for everyone. I’m not saying we shouldn’t take a victory lap before returning to work. However, it’s hard to celebrate when others are falling on hard times, so I’m doing something about it.

Not Every AI Impact Will Be Positive For Everyone Equally

We can’t avoid the darker impacts change will bring to the future of work. It’s an unpopular topic. There are significant risks associated with discussing anything more than what I just did: the new roles and opportunities being created. Someone told me my brand is ‘punk rock’ because I’m willing to start difficult conversations and speak truth to power.

Still, there are topics I’m hesitant to cover symmetrically (the good, the bad, and the ugly), and the future of work is one of them. Before I launch my series on AI agents and platform architecture, I’ll do the responsible thing and explain the implications of this work. It’s a risky conversation that most won’t touch, but more people should speak up. Transparency and plenty of advanced warning are the only ways to safeguard careers.

There’s still time to reskill, but the window is closing. The real punks are the people who keep their mouths shut and watch it happen or profit from it while covering up the impacts to avoid the backlash. The excuse is always some form of "If I’m the one who speaks out, I’m the one who takes the hit. Those people won’t stand with me or pay my bills.” It’s true, but the people profiting must be willing to give back and take a hit.

I watch people hype AI products, knowing businesses will sink billions into vaporware. It’s pathetic. The more time businesses waste on vaporware AI, the more likely they will fail because competitors select better tools. I am all for competition, but businesses failing because they got scammed doesn’t fit my definition of competition. It’ll result in employees being laid off and 401Ks dropping like a rock, all because someone with just enough credibility was willing to make a quick buck and lie about vaporware AI.

There are dozens of legitimate startups and tech incumbents with best-in-class products that get 0 attention because they don’t play the hype game. We miss amazing products daily because we’re forced to wade through influencer slop. Top AI products and some startups will fail all because they had the laughable aim of delivering a pragmatic, high-utility platform that did what its marketing said it did.

NotebookLM is a great example. After I called out Google’s privacy policies, someone from their cloud unit reached out and asked me to give the enterprise licensing another look. I’m evaluating it, and the product continues to find new ways to impress me. While I’m not ready to give it and Google’s Workspace AI suite of tools a full endorsement, I must admit the platform is winning me over.

I no longer use Microsoft’s Copilot, and I spend more time with Gemini than ChatGPT. But you don’t hear honest reviews like that very often.

Jobs Will Be Lost, But It’s The Salary Losses That Will Sting Most

I see the same influencers sweeping the job losses to come under the rug so they don’t offend their corporate sponsors and political connections. People sell courses that teach skills that are being commoditized as we speak. Their students spend thousands of dollars to be trained for redundancy and layoffs.

However, layoffs aren’t the elephant in the room. AI’s impact on jobs was the main topic of discussion during a town hall at Data Day Texas. I heard the words of the influencer set coming out of the audience’s mouths. Influencer slop, much like AI slop, has replaced critical thinking because when you’re overwhelmed by information, you turn to trusted voices for clarity and insight. If hundreds of thousands of people follow this person, they must know something, right?

AI won’t 100% automate a software engineer, data engineer, data analyst, or any other technical role. As I discussed in my last post, AI will make it easier for nontechnical people to deliver technical artifacts. One of the layers in my AI agent platform architecture is a co-development environment where developers and non-developers work together.

Technical developers will build the most complex artifacts for non-technical developers to reuse. There will be more demand for software and technical artifacts, but the millions of nontechnical workers will meet a significant portion of that demand. Technical artifacts are already in a commoditization phase and AI will supercharge it.

What happens next is basic economics. As the value of a work product drops, the value of the skills required to produce it will follow. There is significant demand for restaurant workers, but the amount a restaurant can charge for its products puts a cap on the amount it will pay for workers. Yes, top chefs make top dollar, but most cooks work at the lower end of the pay scale. The increased demand for cooks brought on by the fast food industry didn’t result in massive wage growth.

This is the first trend that most people are sweeping under the rug.

Job Losses Will Come From A Second Driver

Companies are slow to change and will be slow to adopt AI effectively. Conventional wisdom says that legacy companies will need technical roles to maintain their legacy infrastructure. They’ll need people when they eventually modernize. The long adoption cycle will push out the impacts on jobs and mean steady work for several years.

Watch retail this year. Consumer spending isn’t slowing down, but waves of store closures are being announced. This trend will only intensify in 2025 and 2026. We’re watching a consolidation cycle play out, and retail is a sign of things to come.

Companies that don’t transform won’t survive. More mature companies will edge them out of the market. Competition is back, and we will see fewer large companies in 2026 than there are today.

With AI tools and operating models, it takes fewer people to start a business and deliver substitutes for a large incumbent’s products. A smaller, more efficient operational structure allows for lower pricing and products that adapt to customer needs faster. Some large incumbents will slim down and survive. Many will fail.

The incumbent replacement cycle will pour cold water over the slow pace of progress theory. Most legacy code will need no one to maintain it because the businesses that rely on it today won’t be around in 5 years. Inefficient businesses have artificially inflated the demand for technical talent and this unseen bubble is about to burst.

Where The Opportunities Are

I’ll wrap up on a more positive note. There will be new opportunities, and technical skills are still a massive asset. However, the shape of demand is changing. While the skills required to build are being commoditized and salaries are dropping, the domain expertise required to know what to build are seeing the opposite trend.

This week, I asked several people about data and AI strategist and AI product management roles. Do they have a future? Is there a genuine need for them? Almost everyone said they see demand but many had alternative theories of how technical roles would morph into techno-strategic functions.

One person said that “if data architects aren’t capable data strategists, what exactly do they do? It’s part of the job.” Another said that director and VP level technical leaders must play a bigger part in AI strategy. A product manager turned data engineer said more data and AI product managers are building parts of their products. Why wait for the technical team when you can deliver early versions yourself?

Having product and strategy capabilities create opportunities. The future of work isn’t all gloom and doom. I have watched this trend surfacing for over 5 years and it’s why I launched my self-paced courses and certifications. No one else teaches these capabilities so I created an accessible training path to fill the gap.