Three Predictions For 2026
Happy New Year!
As we wrap up 2025, I can't help but think about how quickly things are changing in the AI space.
ChatGPT, for instance, was only created three years ago. Now, it has become an essential part of our personal and professional workflows. CEOs are asking it for business advice and teenagers are copying and pasting text message threads to get dating advice.
Then, there are the fast-moving broader economic effects. While macro analysis is complicated, the rise of AI has undoubtedly affected the job market. As a recent Business Insider report argued, submitting a job today is like shouting into the void. Applicants are using AI to draft their cover letters and edit their resumes. Employers are using AI to filter through those applications. Consequently, applicants submitting online applications today have around a 0.4% chance of actually getting the job.
As the technology gets better, it isn't controversial to expect these trends to accelerate. We're not even close to seeing some of the significant disruption (both economic and otherwise) that almost seems inevitable at this point.
It's hard to make predictions about such fast-changing technology. But to wrap up this year (and to have a little fun), I wanted to share three quick predictions about the AI space next year. I may be totally off with some of these, but regardless, I'm excited to revisit them next December!
First, judgment and taste are the new sources of edge. As AI continues to get better at basic (and more complex) skills, the actual skills themselves will become more like a commodity. What will distinguish humans (and digital products) are the grey areas where those skills reside. Because context windows will remain constrained, humans can provide the most value by handling edge cases. From white collar work to travel planning, we need to lean into nuance and focus on situations and circumstances that models still can't replicate.
Second, agent marketplaces are going to become much more popular. There has been plenty of buzz about the power of agents throughout the year. In the past few days, there has been tons of enthusiasm about Claude Code and how the average person can manage a team of agents to complete all kinds of tasks. At the same time, few people want to create agents themselves. I think that agent marketplaces (one good example is from HubSpot) will be used to complete more complex digital work. Companies (and even individuals) will "hire" agents to complete specific tasks. Like people, the "best performing" agents will cost more than those that are less effective.
Finally, backlash against data center growth and AI-related job loss is going to play a larger-than-expected role in the midterm elections. I've previously written about the growing pushback against data center construction throughout the country. I'd expect more of this as we get closer to November. And if the economy does go into recession this year, I'd expect a good number of politicians (on both sides of the aisle) to place blame on AI.
Again, I could be wrong! But regardless, it's going to be a fascinating year. I wish you the happiest of new years and an excellent start to 2026.
Prompt of the Week
It's New Year's Resolution time. But instead of helping craft specific resolutions, I asked ChatGPT to help keep me accountable. The worst thing (most of us are guilty of this) is spending time creating resolutions to eventually forget about them. To avoid this problem, try this prompt right after describing your resolutions.
"I want you to think hard and develop a plan to help me execute my New Year's resolutions. When developing a plan, be as practical as possible. Consider my schedule and work demands. Accomplishing even some (or part) of my resolutions is substantially better than having a super ambitious plan and accomplishing none of them at all."
Happy New Year!