I’ve been around the tech industry since 2000, and if there’s one thing I’ve learned, it’s that every new wave of technology comes with a lot of noise. What’s clear now is that the world of artificial intelligence is at an inflection point. The hype curve is real with everyone talking about AI and agents that can actually do work for you, and whether we’re in some kind of AI bubble.
Critics point to the high failure rate of new initiatives. But beneath the surface, a sea of change is underway that is reshaping productivity, the workweek, and the broader economy. This may be the case that the bubble doesn’t pop.
Why We’re Not in a Bubble
It’s easy to draw parallels between today’s AI boom and previous tech bubbles. As was recently widely reported, 95 percent of AI initiatives fail because they don’t get traction, they don’t deliver ROI, and most of them just fade away. But that’s not a sign of a bubble. That’s a sign of rapid experimentation and a market that’s finally able to sort itself out.
Back in the day, if you wanted to switch tools, it was a huge lift—long sales cycles, big integration projects, lots of inertia. Now? If something doesn’t work, people drop it instantly and move on. The switching costs are basically zero.
What we’re seeing is the end of the “imagination phase.” People aren’t just playing with AI for fun anymore. They want real results: actual time saved, workflows automated, and productivity that you can measure. If a tool doesn’t deliver, it’s gone. That’s not a bubble. That’s evolution. The winners in this new era will be those who move beyond demos and deliver real, sustained impact.
What Agentic Tech Really Means
Agentic technology is more than just a new buzzword. It represents a fundamental shift in how software interacts with people and organizations. Instead of isolated tools that require constant human input, agents can automate entire workflows, retain organizational knowledge, and act proactively on behalf of users.
So what’s different about this wave? It’s not just about having an agentic AI assistant everywhere, though that’s still a big part of it. The real shift is toward what I call a “storage of intelligence.” Think about all the knowledge employees create: meeting notes, project documents, best practices. When an employee leaves, it’s tribal knowledge that leaders fear losing. They worry about what knowledge walks out the door with them. But what if you could capture it, store it, and actually use it across the company, even after they’re gone?
We’re also starting to see the rise of what I call the “insurance of intelligence.” If you have a key player in your organization, how do you back them up? Not just their files, but their actual decision-making, their expertise. We’re not far from having digital twins—AI agents that can answer questions and make decisions just like that person would, because they’ve absorbed all their work history and communications. This isn’t a face-filter that gets a laugh from your friends, it’s a real change in the way companies store intelligence.
Impact on Productivity and the Workweek
One of the most visible effects of AI agents is the transformation of the workweek. New research shows that for AI-empowered teams, meetings are fewer and moving to the middle of the week. Mondays and Fridays are becoming full-focus days—heads-down, no-meeting time. AI is helping people optimize, moving things around, and so workers can get more real work done in a cadence that works better for their needs.
Here’s the thing: not all agents are created equal. Tools that promise the world but don’t actually automate a specific workflow fails. Solving a single workflow is valuable, solving all workflows is vaporware in today’s market. Some of the best-known copilots are great examples. Tons of traction, but adoption hits a wall because the platform doesn’t automate end-to-end; it stays within its own silo. The real winners are the agents that act independently, automate the work no one wants to do, and let people focus on what matters.
To be clear: Having your team build agents doesn’t make sense. The best next step is to understand what your team is doing manually today, and the workflows they can automate. Then identify agents pre-built to solve these problems and rinse and repeat. Ideally you pick a partner that has the ability to orchestrate these agents, not just for the individual or team, but for the entire organization, enabling multiplayer synergies and exponential growth.
Economic and Organizational Implications
This isn’t just about individual productivity. The economic impact is massive. Agentic tech compresses adoption timelines: what used to take quarters now happens in weeks and days. Organizations can scale faster, adapt quicker, and the old barriers (slow procurement, complex integrations) are disappearing.
Here’s what really drives the point home: A recent study found that publicly traded companies adopting productivity AI are outpacing the broader market. These companies saw revenue gains of more than 2X, with an average year-over-year increase of 13%. By comparison, the S&P 500’s index-weighted average growth over the same period was under 6%. It’s clear that AI-driven productivity isn’t just a buzzword; it’s delivering real, measurable results.
Another shift: how these tools get adopted. In the old days, it was all top-down. The CIO picked a tool, and everyone had to use it. Now? It’s bottoms up. Developers, marketers, whomever—they find something that works, start using it, and it takes hold. That’s how Slack won. That’s how the best AI tools are winning now. If you’re relying on top-down mandates, you’re already behind.
Just like employees bought computers to work from home and smartphones to check emails before it was required, employees are self-selecting into using AI, and employers need to amplify that adoption. To limit the adoption of AI in your organization is to promote a grey market of AI tools.
Addressing Skepticism and Looking Forward
Skepticism about AI is healthy, even necessary. We need to talk about governance, ethics, and what happens when agents go mainstream. But the high failure rate isn’t a sign of a bubble. It’s a sign that the market is working, bad solutions fail, and great solutions win and drive more adoption within an organization.
Enterprises need to avoid the ‘grey market’ of AI adoption and establish proactive policies around AI adoptions, and qualifying vendors. Questions to ask to get ahead of it all: Who will decide how agents are used at our organization? How is success measured? What safeguards are in place to ensure ethical deployment? How do we communicate and implement our commitment to privacy and transparency? These questions will define the next phase of agentic tech and determine which companies lead the way.
What’s Next:
We are in a bubble, but this time it’s different, where adoption is mainstream, where revenue is real, and the impact is tangible. Fearing when the bubble will pop, shouldn’t be at the expense of what this Cambrian explosion of AI will have on the global economy.
We’re in a period of rapid evolution. The winners will be those who focus on utility, interoperability, and ethical deployment. The future of work is agents that augment human capability, drive economic growth, and empower us to focus on what we do best. Agentic AI is already reshaping our world, and the gap between the haves and have-nots will continue to widen.
