If AGI Arrives In 2027, What Should A Local Business Do This Year?

Mo Gawdat, formerly Chief Business Officer at Google X, has put a number on it: artificial general intelligence by 2026 or 2027, then roughly twelve to fifteen hard years before things get better.

I want to handle this the way I would want somebody to handle it for me. Not dismissed, because of who is saying it and how specific it is. Not adopted as fact either, because nobody predicts a decade of technology correctly, and the graveyard of confident forecasts is very full.

Here is what I think a business in this valley should actually do with a claim like that.

The test for any long forecast

Ask what you would do differently if it were true, and then ask whether that action is also correct if it is false.

If the answer to the second question is yes, do it now and stop thinking about the forecast. If the answer is no, you are being asked to place a bet, and you should size it like a bet rather than a plan.

Almost everything worth doing in response to this prediction passes the first test, which is convenient and also the reason I am not especially worked up about it.

The moves that are correct either way

Build capability, not dependence. Learn what these systems do, where they fail, and how to check them. That knowledge is valuable if AGI arrives in 2027 and equally valuable if it arrives in 2045 or never. A subscription you do not understand is worthless in every scenario.

Keep your workflows portable. Do not construct your operation around one vendor's specific features. If the landscape shifts hard, portability is what lets you move. If it does not shift, portability costs you almost nothing.

Deepen relationships. In the scenario where machines do most cognitive work, the scarce thing is trust in a specific person. In the scenario where nothing much changes, relationships were always the moat for a local business. There is no version of the future where knowing your customers hurts you.

Stay financially flexible. Twelve to fifteen hard years, if it happened, would mean volatility rather than a single event. Low fixed costs and short commitments are how a small business survives volatility, which was true before anyone said AGI.

Every one of those is ordinary good practice. That is the point.

What I would not do

Reorganize the business around a timeline nobody can verify.

I have watched people make genuinely bad decisions on forecasts like this. Delaying a hire that the business needed now. Skipping an investment with a clear three year payback because of a prediction about year five. Buying a large software commitment because a vendor attached it to an inevitability.

The prediction might be right. It is still not specific enough to act on, because it does not tell you what happens to your industry, in your market, on what schedule. A forecast you cannot convert into a dated action is information, not instruction.

The part of the prediction I take most seriously

Not the AGI date. The twelve to fifteen hard years afterward.

That framing is unusual and I think it is the useful part. Most predictions in this space are either utopia or apocalypse, and both are easy to ignore. A forecast that says the technology arrives and then there is a long, difficult, messy adjustment sounds much more like how every major technology has actually gone.

If you want one thing to plan around, plan around adjustment being long and uneven rather than a sudden switch. That means resilience matters more than positioning, and the ability to change your mind matters more than picking correctly right now.

What I am doing

Same thing I would tell you. Building systems I understand, keeping them portable, spending more time with people than with tools, and not making any commitment that only pays off if a specific version of the next five years shows up.

If AGI arrives in 2027, that posture is right. If it does not, that posture is still right. That is the only kind of plan worth having about a decade you cannot see.

Common questions

Connor T. MacIvor · CalDRE #01238257 · Sync Brokerage, Inc. · DRE #02031490