Message

I do independent AI safety research. If you want to hire, mentor, make me your fellow, or simply be friends, DM me or email xjohanbonilla@gmail.com.

3mo

Johan David Bonilla

I do independent AI safety research. If you want to hire, mentor, make me your fellow, or simply be friends, DM me or email xjohanbonilla@gmail.com.

Johan David Bonilla — LessWrong

Johan David Bonilla

Message

I do independent AI safety research. If you want to hire, mentor, make me your fellow, or simply be friends, DM me or email xjohanbonilla@gmail.com.

3mo

Johan David Bonilla

I do independent AI safety research. If you want to hire, mentor, make me your fellow, or simply be friends, DM me or email xjohanbonilla@gmail.com.

Mathematics After AI

Mathematics is experiencing the same generative boom that software engineering and many other fields have been undergoing over the past few years. A lot of the reaction has been closer to mourning than excitement. That reaction is understandable when people have spent their lives mastering a craft that suddenly looks...

Aug 4•1

The Name is Not The Model

Abstract A safety evaluation has to trust something to know what it is testing: the model's name, its version string, and the reasons the model gives for what it does. On one deployed alias, none of the three held. I sent the same 100 harmful requests to gemini-3.1-pro-preview through two...

Jun 30•2

Johan David Bonilla's Shortform

Jun 7•1