AI’s Future In 2027: 10 Things I Expect, With Confidence Ratings

🔍 Read the full analysis: AI’s Future In 2027: 10 Things I Expect, With Confidence Ratings on ThorstenMeyerAI.com

Buying for a business?Offer from Amazon

Get business pricing on office and shipping supplies

• Business-only prices and quantity discounts
• Tax-exempt purchasing
• Multiple users, one account, clear invoices

Create a free Business account →Free to register

As an affiliate, we earn on qualifying purchases.Hide

Thorsten Meyer published 10 predictions for AI in 2027 on October 15, rating each by his confidence and outlining evidence he says will be used to assess the calls by October next year. The forecast argues that AI use could shift from one-off prompts toward delegated work, while identifying verification, data control and electricity supply as important constraints; these are Meyer’s expectations, not confirmed outcomes.

Meyer says his 10 confidence ratings average 72.5%, which he interprets as a prediction that about seven calls will prove correct if his confidence is well calibrated. He characterizes the piece as a forecast rather than a prophecy and says each prediction includes signals for judging whether it came true. The provided source text does not include all 10 predictions or all of their scoring criteria.

Among the five predictions described in full, Meyer gives an 85% confidence rating to the idea that AI assistants will retain information about users’ work between conversations. He assigns 80% to the view that evidence and verification will become more important as AI-generated answers grow cheaper, and another 80% to the prediction that energy, chips and cooling will constrain AI capacity. He expects routine tasks to move partly onto personal devices if cloud computing remains constrained.

Meyer also predicts that people will delegate work lasting several days to AI agents, a call rated 70%, and that most new code will be produced by AI rather than typed line by line by people, rated 75%. A sixth prediction, rated 65%, begins with AI agents acting as research partners by proposing approaches and running simulations. The supplied text ends before that prediction is complete and does not provide predictions seven through 10.

How Meyer Expects AI Work to Change

The forecast focuses on a shift in the human role: from repeatedly prompting a system toward setting goals, defining acceptable results and checking work. If agents take on longer tasks, the practical test will not simply be whether they can generate an answer or code, but whether they can stay on task and show how they reached a result.

Meyer’s emphasis on verification reflects a key risk in relying on automated output. A plausible report or working program may still be wrong. His predictions point to greater demand for tests, records of actions and independent review before AI-generated work is accepted. The claims are forecasts; the article does not present independent measurements showing that these changes will happen at the predicted scale.

The predictions also connect AI adoption to infrastructure and control. Persistent assistant memory could make tools more useful across projects, but it would also hold sensitive information about a user’s work. At the same time, limits on power and computing capacity could affect cost, availability and whether tasks run in the cloud or on personal devices.

The Forecast’s Ratings and Scope

Meyer published the article on ThorstenMeyerAI.com on October 15, 2026. He says confidence ratings are intended to make predictions testable: a rating of 70%, in his explanation, means he would bet on the outcome but expects to lose that bet roughly three times in 10. He says the average across his 10 calls is 72.5% and sets October 2027 as the point by which readers can judge them.

The first five calls cover multi-day AI agents, AI-generated code, verification, assistant memory and electricity constraints. The sixth, only partly included in the available source, concerns AI supporting scientific research. Meyer describes possible systems proposing research approaches, running simulated experiments and sending promising candidates for proof checks. He also cautions that faster generation of hypotheses could leave testing and human judgment as bottlenecks.

The article references Meyer’s earlier writing on AI development and an account of building games with AI, and links to a video summarizing the predictions. Those references provide his perspective, but the supplied material does not cite outside studies or present corroborating data for the 2027 forecasts.

“A forecast, not a prophecy.”

— Thorsten Meyer

What the Predictions Cannot Confirm

The predictions describe what Meyer expects, not what has already happened or what independent evidence establishes will happen in 2027. The source text provided for this report is incomplete: it stops partway through prediction six and does not state the content, ratings or evidence criteria for predictions seven through 10. The article’s stated average rating covers all 10, but the missing entries cannot be checked against that figure here.

It is also not yet clear how Meyer will define success for broad claims such as “most new code” being produced by AI or routine AI work moving onto personal devices. The supplied text refers to scoring signals under each prediction, but those detailed signals are not included. No independent assessment of the ratings or the forecasts is provided.

Scoring the Calls by October

Meyer says he intends to assess the predictions by October 2027, using the signals he says accompany each call. That assessment should show which forecasts he counts as successful and how he applies the ratings; the supplied source does not specify a separate review date or an independent scoring process.

Readers would need the complete list, including predictions seven through 10 and the full scoring criteria, to evaluate the 72.5% average against all of Meyer’s claims. Until those details and the review are available, the predictions remain forward-looking judgments, not evidence that AI adoption will follow this path.

Key Questions

Who made the 2027 AI predictions?

Thorsten Meyer published them on ThorstenMeyerAI.com on October 15, 2026.

What does the 72.5% figure mean?

Meyer says it is the average confidence rating across his 10 predictions. He interprets the average as implying that about seven calls should be correct by October 2027 if his ratings are well calibrated; it is his forecast metric, not an independently verified probability.

Which predictions are described in the supplied article text?

The source gives five predictions in full: multi-day AI agents, AI-generated code, greater emphasis on verification, persistent assistant memory and energy constraints. It begins a sixth prediction about AI as a research partner but cuts off before completing it.

Are these outcomes confirmed?

No. They are Meyer’s expectations for 2027. The source does not establish that the predictions have occurred or provide outside evidence confirming them.

Source: ThorstenMeyerAI.com