Every year produces a list of technologies that will change everything, and every year most of them do not. The problem is not that the predictions are dishonest. It is that “emerging” describes two completely different situations that get written about identically.
Some technologies are early but working — the constraint is adoption. Others are impressive but unsolved — the constraint is a hard technical problem nobody has cracked. Telling them apart is the whole skill, and it is not difficult once you know what to ask.
Four questions that separate ready from hyped
- Is anyone boring using it in production? Not a startup demo — a bank, a logistics firm, a government department. Boring organisations adopt things after they work, which makes them an excellent signal.
- What is the failure mode when it goes wrong? Technologies with graceful failure get adopted early. Technologies that fail catastrophically wait for regulation and insurance, which takes years.
- Does it need everyone else to adopt it too? Anything requiring network effects moves at the speed of the slowest participant. That is a decade, not a year.
- Has the cost curve actually bent? Many technologies work but cost ten times the thing they replace. Watch the price, not the capability.
Applying it to what is on the list this year
| Technology | Boring users? | Failure mode | Needs others? | Verdict |
|---|---|---|---|---|
| Practical AI assistants | Yes, widely | Graceful — wrong answer | No | Ready now |
| Document extraction | Yes — insurers, accountants | Graceful — flagged for review | No | Ready now |
| Edge computing | Yes — retail, industrial | Contained | No | Ready, unglamorous |
| Autonomous vehicles | Limited pilots | Catastrophic | Partly — regulation | Years away for most |
| Quantum computing | Research only | N/A | No | Not a business decision yet |
| Digital currency rails | Some | Financial | Yes, heavily | Depends entirely on regulators |
| AR in the workplace | Narrow niches | Graceful | No | Real but small |
The two that matter for most businesses
Document extraction. Unglamorous, well past the demo stage, and it removes a task nobody wants. If your business types numbers from PDFs into spreadsheets, this is the highest-return technology available to you and it will not appear on any keynote stage.
Assistants for verifiable tasks. Transcription, summarising, drafting, formulas. Ready, cheap, and useful precisely where mistakes are obvious. We put actual numbers to that in our six-task timing test.
Neither is exciting. Both are working today, for ordinary companies, at prices that make sense. That combination is what “emerging” should mean and rarely does.
Why hype cycles repeat
Three forces keep producing the same pattern, and none of them is going away:
- Demos are the easy 80%. A working demo genuinely is a real achievement — it just is not evidence about the remaining 20%, which is where the years go.
- Investment needs a story. Capital flows toward narrative, and narrative rewards big claims over calibrated ones.
- Nobody is rewarded for correct scepticism. Being right that something would take a decade earns nothing. Being early and loud earns a great deal.
Understanding this is not cynicism. It is the reason to run the four questions yourself rather than reading a list of predictions, including this one.
How to evaluate without gambling
| Stage | Time to spend | Money to spend | Decision to make |
|---|---|---|---|
| Watching | An hour a quarter | Nothing | Is this getting cheaper? |
| Small test | A few days | Under a month’s licence | Does it work on our real data? |
| Pilot | A month | One team’s budget | Does anyone actually use it? |
| Adopt | Ongoing | Real budget | What do we stop doing? |
The last column is the one people miss. Adopting a technology without stopping something else means you now maintain both, which is how organisations accumulate tools nobody has owned since the person who bought them left.
Common questions
Is being an early adopter an advantage?
Sometimes, in narrow windows, for businesses where the technology is core rather than supporting. For everyone else, being an early second adopter — letting someone else find the sharp edges — is a better trade.
How do I keep up without wasting time?
An hour a quarter, spent asking the four questions about anything you keep hearing about. That is enough to catch a genuine shift and cheap enough to sustain.
What if a competitor adopts first?
Watch what happens to them. Most first movers on unready technology spend a year discovering limits, and you get that information free.
Does this apply to buying software generally?
Yes, and the same instinct — look for specifics, distrust balanced enthusiasm — is what separates a real review from marketing. We wrote that up in how to read a software review.
Leave a Reply