ai predictions

People on the internet make big predictions about AI.

Many of these claims are dramatic. Some border on the hysterical. It’s difficult to tell the true believers from the liars and frauds trying to grab attention. The purpose of this post isn’t to critique these claims, but to record them for the purpose of evaluating them later.

Posting of this type finds a natural home on Ecks the Everything (And We Really Mean Anything) App. I chose seven examples that make predictions with specific timeframes. The posts are from accounts with large follower counts that are representative of the claims often made or boosted by accounts that involved with this topic.

I think that the predictions are incorrect, but the ephemerality of posting means these claims slip away and are forgotten – and perhaps deleted – as the vibe shifts over time. So I’ve taken screenshots – I’ll update this post periodically once enough time has passed to check each claim.

If you have any more examples, please send them to me and I can include them here.

@davidpattersonx

https://x.com/davidpattersonx/status/2008429541417787442

I’m going to read this as it being possible to generate games/movies that are equivalent to a traditional movie or game (1-2 hours of screen time for a movie, multiple hours of playtime for game) and that people are really watching or playing the results – not just showing a 3 minute tech demo.

@iruletheworldmo

https://x.com/iruletheworldmo/status/2007987156690805052

This prediction isn’t specific, so it will be hard to evaluate objectively. I’m going to reduce scope a little and assume the stars and planets will continue to move through the heavens as expected in 2026. Let’s assume “the universe” just means Earth.

To say this has come true, I’d want to see significant changes to national borders, new social and political movements resulting in real changes to existing systems, and revolutionary advances in technology and scientific understanding.

@EMostaque

https://x.com/EMostaque/status/2007513768750919840

This post doesn’t explicitly refer to AI, but Demis Hassabis (in the quoted post) says “I realized in the last 20 or 30 years, we haven’t made much progress in understanding some of these fundamental laws. So I thought – why not build the ultimate tool to help us – which is artificial intelligence.” I think it’s reasonable to infer Emad is saying AI will be the tool that gives us a “very cool” unified theory. I’m going to take this as a prediction that we’ll have a proposed unified theory which is testable and widely accepted by the physics community by the end of 2027.

@beffjezos

https://x.com/beffjezos/status/2009197776744173583

I cheated a bit here as the poster isn’t really making a claim, but @beffjezos posted a clip from an interview with Elon Musk where he claims “[In three years] there will probably be more Optimus robots that are great surgeons than there are all surgeons on Earth. If you put a little margin on it, better than any human in four years.” According to World Health Organisation data, there are 953496 active surgeons worldwide. The data seems to be from the mid-2010s, so I’ll round that to a million.

So for this to be true, I’d need to see a million deployed Optimus robots performing surgery on humans and be widely accepted by the medical community as superior to a human equivalent, by 2030.

@DeryaTR_

https://x.com/DeryaTR_/status/2006821246147899832

How we decide what represents a century’s worth of progress is subjective. I read Derya as saying we’ll get 25 years of progress from 2026 to 2030, and then another 75 years by 2035, packing a full century into the next decade. Like @iruletheworldmo’s post, this is subjective.

For me to agree this prediction was correct, I’d want to see revolutionary advances in communication, transportation, energy, and health by 2035, equivalent to the transformations in those areas during the 20th century. I’m not sure how to quantify 25 years of progress by 2030; I’ll call it how I see it in 2030.

@IterIntellectus

https://x.com/IterIntellectus/status/2009605453073072422

This one is fairly straightforward. We should expect to be able to talk to animals with our brains by 2036 for this to be true.

@Dr_Singularity

https://x.com/Dr_Singularity/status/2008694311181750419

@Dr_Singularity didn’t hold back for this one. Luckily, again, these predictions are quite straightforward. We can check if this is true by seeing if there are invisible forcefields, control of gravity, and immunity to axes by the end of the 2030s.