On: Introducing deep research

[tmnvdb]

Eating popcorn while the scaling doubters scramble to move the goalposts for the nth time.

[elicksaur → tmnvdb]

Its number for one of the benchmark has:

**with browsing + python tools

Maybe we have different definitions of scaling?

[tmnvdb → elicksaur]

I would consider unsupervised tool usage an achievement

[elicksaur → tmnvdb]

But it’s not simply scaling. Who is moving the goalposts exactly?

[khazhoux → tmnvdb]

Trying to parse this. What are you saying?