Will LLMs become a ubiquitous part of everyday life by June 2026?

Ṁ1kṀ4.1k

Jun 30

91%

chance

ALL

On par with current google search usage or current smart phone usage. If most people use a LLM in any capacity at least once per day market resolves to YES. subject to my judgement. I won't participate in this market

To clarify "by" in this context means whether this will be true at the close date. I won't resolve before then.

Market context

Technology

Get

1,000

to start trading!

People are also trading

At the beginning of 2028, will LLMs still make egregious common-sensical errors?

64% chance

Will there by a major breakthrough in LLM continual learning before 2027?

43% chance

Will LLMs Daydream by EOY 2026?

20% chance

Will we have a popular LLM fine-tuned on people's personal texts by June 1, 2026?

45% chance

Will an LLM improve its own ability along some important metric well beyond the best trained LLMs before 2026?

14% chance

Will a frontier-level diffusion LLM exist by 2028?

30% chance

Will the most advanced LLM stop being from a US-based company any time before 2030?

34% chance

Will I write an academic paper using an LLM by 2030?

65% chance

Will LLMs be used for academic peer review by 2030?

71% chance

LLMs widely used in economics modeling by the end of 2026?

Sort by:

https://bsky.app/profile/miaow.ing/post/3m7urjqvevc2x

getting frequent phone calls from LLMs by end of 2025 is pretty high bar ngl, wouldn't have expected that early 2023 ahaha

Google search gives me an LLM-generated result 30-50 percent of the time on the front page. Combine this with the popularity of ChatGPT and other assistants, LLM-generated suggestions on many websites, google translate using an LLM and other mentiones below, I'd say it's safe to resolve this Yes by now.

@ProjectVictory As per description, the market can not resolve before close date

They don’t know that google search and autocomplete on every modern phone is just LLMs

@Gigacasting Also all translation - page translate, google translate, translation on social media, autogenerated youtube etc. Maybe you are going to argue that the autosuggestion model running on iphone keyboards isn't "large", but I would think that the models used in translation are considered LLM by most.

I argue that this should probably be resolved or clarified, because right now I assert that a majority of people see and interact with LLM on a daily basis.

@floomby To clarify "by" in this context means whether this will be true at the close date. I won't resolve before then.

@therealfloomby Is the autocomplete model a transformer model, with attention and KV and tokens? Or some other ML setup?

@EvanDaniel As of IOS 17 in 2023, it is based on transformers. It wasn’t when this market was created 👀

How do you plan to assess it? Google trends suggest it's very far although I'm not sure how well it reflects usage. I'd guess it does very well though.

@na_pewno simply by observing the world around me. Probably not using any kind of data

@DylanSlagh And I assume "the world around you" is disproportionately filled with early adopter tech people?

@jonsimon not particularly. On social media yes admittedly. As I said in the description, on par with search engine usage and smart phone usage. 80% of people in the US own a smart phone. So that would be a good benchmark. So it's a very high bar. I think the most likely YES scenario is if every google search result contains output from a LLM