Ask HN: Effective Methods for Predicting Technology?

3 points by wildermuthn ↗ HN
A recent interview with the CTO of OpenAI, Greg Bockman, made a fascinating point: new technology is hard to predict (“imagine predicting Uber and smartphones in the 50s”).

And yet good science fiction writers routinely do just that. I think in particular of Forster’s “The Machine Stops.”

What are effective methods for predicting future technology?

2 comments

[ 2.7 ms ] story [ 15.8 ms ] thread
> yet good science fiction writers routinely do just that.

I see two things going on here.

- Science fiction influencing engineers / scientists.

- Selection bias, how many sci-fo books/movies are completely wrong ? If you produce 100 000 scenarios surely some of them will be more or less comparable to reality.

Did back to the future predict self lacing shoes ? Or did Nike invent self lacing shoes because of back to the future ? &c.

And yet good science fiction writers routinely do just that. I think in particular of Forster’s “The Machine Stops.”

This is survivorship bias. Between them all scifi authors predict everything they can imagine, and a few of them happen to get close enough to what happens that they look accurate, and people conveniently forget all the ones who were wildly off the mark.