I'm struggling to keep up with AI. There, I said it.
I’ve been researching and writing about AI since 2011; run an AI-native education company; host a weekly AI podcast; speak about AI at dozens of conferences and private events every year; experiment daily with the latest AI tools; and consume hundreds of articles, reports, podcasts, videos, and social posts about AI every week.
Yet, I continually feel like it’s not enough.
This has been exacerbated in 2026 by the emergence of more reliable—and increasingly autonomous—AI agents (i.e., AI systems that take actions to achieve a goal).
I hit a bit of a personal breaking point last week after attending Google Cloud Next in Las Vegas.
By my flight home on Wednesday, we already had major AI agent news from Adobe, Google, Microsoft, and OpenAI:
Then OpenAI introduced GPT-5.5 on Thursday, April 23.
So in addition to feeling like I’m not fully grasping the opportunity and impact of the agentic ecosystem—Claude Code and Cowork, OpenAI Codex, OpenClaw, Lovable, Replit, etc.—that is putting the power of coding in the hands of mere mortals like me, I have to figure out which agent platform we should build SmarterX around.
Should we grab some Mac Minis and accept the risks of experimenting with emerging technologies like OpenClaw?
Should we push on vibecoding apps and tools with Lovable and Replit, and figure out how to move them safely into production and public release?
Should we double down on ChatGPT, giving OpenAI access to all our data and infusing their agents into our workflows?
Should we expand our use of Anthropic Claude and train everyone on Cowork to transform how we scale?
Should we bet on Google Cloud and their unmatched AI infrastructure and financial strength?
Or should we keep spreading our bets, never fully maximizing the impact and returns of any one platform?
I don’t know if there is a right answer.
The space is too dynamic, and the different models and companies bring unique strengths and capabilities to the market.
But, I needed to do something to start figuring this out.
So, on the flight home from Vegas, I decided to create a new internal business unit called SmarterX Labs, and I scheduled our first “Vibe” for the following week. (Yes, I know the Vibe name is completely unoriginal, but it just seemed to fit).
I defined Vibes as builder sessions (for non-coders) designed to rapidly prototype apps and agents that optimize workflows, solve business problems, and accelerate growth through innovation.
Vibe 1, which I kept intentionally to a very small group to start, was focused on the new ChatGPT agents. What are they? How could we use them? What could we build in under an hour as a proof of concept?
Ironically, in the process, we realized that Gemini Enterprise may be the better solution for us, but I’ll save that story for a future podcast.
I guess my point with all of this is that if you’re feeling overwhelmed or like you’re falling behind and not doing enough, you’re not alone.
My only advice is to just take the next step.
We’ll do our best to keep sharing what we’re trying and learning at SmarterX, and we’ll continue to invest in creating a community of AI-forward professionals and leaders who can help each other through our personal and business AI transformations.
Want This Content Sooner?
You can gain access to these insights before they appear on our website! Subscribe to Paul’s ExecAI Newsletter and get it delivered to your inbox every Sunday to start your week.