# If you think that scaling the current paradigm won’t get us to AGI, where does it fail? — retained record

> Retained reddit record from reddit:singularity.

Canonical: https://highsignal.app/data/records/35340ca9d0de18aa

[← sources](https://highsignal.app/data)


source: [reddit](https://highsignal.app/data/reddit) (reddit:singularity)  published 2026-09-07 17:15 ingested 2026-09-08 02:39

[original source ↗](https://www.reddit.com/r/singularity/comments/1w9y6l6/if_you_think_that_scaling_the_current_paradigm/)

## retained text



```
The bet from OpenAI and Anthropic seems pretty straightforward: keep improving their LLMs/reasoning models, memory, maths, coding, tool use and agents until you get a genuinely capable automated AI researcher. That researcher then helps accelerate AI research, producing better models and, in turn, a better AI researcher.

Today’s LLM architecture doesn’t need to take us all the way to ASI. It only needs to get us to the point where AI can meaningfully automate AI research and help develop whatever comes next.

If you think that loop breaks down, where exactly does it break?

And what evidence would make you change your mind?
```



document fetched 2026-09-08 00:31 · canonical https://reddit.com/r/singularity/comments/1w9y6l6/if_you_think_that_scaling_the_current_paradigm
