When I suggest everyone is going faster than you, the inevitable question I get back is “but where are they going?”
I have always been of the mind that fast beats right, so my pithy answer is “forward,” but I'll give you a few thoughts on how to apply all those subsidized LLM tokens productively.
There are three places I think increased building capacity should go:
Things you should have been doing, but weren't
This is the simplest place to start, which is to use AI to do the things you are already doing, but do them well and do them completely. This means no more “managing scope creep” -- build those previously cut features. Invest in design polish. Burn down your tech debt. Execute in the marketing channels that were never a budget or headcount priority, but that you suspect are leaving money on the table. Localize. Speed up. Gamify. Experiment.
You have a roadmap your roadmap, now execute it as if you have a 10x the team investing against it. You know the list, it's “below the cutline” or a “FY27 thing.” Write it down. Build it today.
Internal tools
Finally, the revenge of the internal platform team. Previously, only product or customer facing building held cachet. If you were “backoffice” or “platform” or the dully named “internal tools” you were starved of budget, headcount, and internal prestige.
No more.
Right now companies should be investing more in these teams because a) the cost of investment has dropped and b) the leverage in these investments is higher than it's ever been.
This is even more applicable in B2B, where everyone will yell “our customers and the market and oh god the FIELD cannot absorb all these yolo'd features!” You do not have to point your AI building capacity exclusively at externally facing customer innovation. You can build dashboards, customs ops tooling, and in EPD, invest in DevX and CI/CD or a custom background agent.
Put your best builders and point them inside the house. It might be the best investment you can make in the next 12 months.
Non-linear bets