#### Your submission is now in Draft mode.

Once it's ready, please submit your draft for review by our team of Community Moderators. Thank you!

#### Your essay is now in Draft mode

Once you submit your essay, it will be available to judges for review and you can no longer edit it. Please make sure to review eligibility criteria before submitting. Thank you!

Submit Essay

Once you submit your essay, you can no longer edit it.

#### Pending

This content now needs to be approved by community moderators.

#### Submitted

This essay was submitted and is waiting for review by judges.

# How many FLOPS for human-level AGI?

### Question

A major uncertainty in understanding some timeline estimates for high-level AI is in estimating the minimal computational power necessary to perform the operations that the human brain does.

Estimates in the literature (see appendix A on p. 84 of this paper for a compilation) range from $10^{12}$ to $10^{25}$ FLOPS following a variety of methodologies. (For a comparison, the Landauer limit at 20 C is about $10^{22}$ bit erasures per second. However, the author has no clear idea how to convert between bit erasures and FLOPS.)

This huge range will probably eventually be narrowed down to within an order-of-magnitude or two, and we ask for that number here. Assume that by 2075 there is either (a) a full software emulation of a human brain that can duplicate the basic functionality of a typical adult human of average intelligence; or (b) there is an AI system that can pass a full "strong" Turing test (i.e. the interview is long, adversarial, and include sensory data); or (c) there is a computer system that attains "human intelligence parity" by the definition set forth in this Metaculus question. In each case (a,b,c) the number will be evaluated on a state-of-the-art system five years after the first demonstration of a system satisfying the criterion.

What will the computation in FLOPS be of this machine system, if running at a speed comparable to that of human mental processing?

The point of this question is not really as a prediction, but more as a gathering place for estimates.

Fine print: we'll settle for a published estimate accurate to within a factor of 5. The speeds of the systems can be matched up by requiring that similar delays occur between queries and responses in the system as compared to humans, or scaling for this equivalency. Resolves as ambiguous if (a), (b) or (c) don't occur by 2075.

(edited 2020-09-13 to fix eval date as 5 years after such a system appears.)

### Prediction

Note: this question resolved before its original close time. All of your predictions came after the resolution, so you did not gain (or lose) any points for it.

Note: this question resolved before its original close time. You earned points up until the question resolution, but not afterwards.

This question is not yet open for predictions.

Current points depend on your prediction, the community's prediction, and the result. Your total earned points are averaged over the lifetime of the question, so predict early to get as many points as possible! See the FAQ.

### Metaculus help: Predicting

Predictions are the heart of Metaculus. Predicting is how you contribute to the wisdom of the crowd, and how you earn points and build up your personal Metaculus track record.

The basics of predicting are very simple: move the slider to best match the likelihood of the outcome, and click predict. You can predict as often as you want, and you're encouraged to change your mind when new information becomes available.

The displayed score is split into current points and total points. Current points show how much your prediction is worth now, whereas total points show the combined worth of all of your predictions over the lifetime of the question. The scoring details are available on the FAQ.

#### Thanks for predicting!

Your prediction has been recorded anonymously.

Want to track your predictions, earn points, and hone your forecasting skills? Create an account today!