HNHacker News
TopNewBestAskShowJobs

SebastianSosa1

2 karma · joined April 28, 2025

submissionscomments
SebastianSosa1··on 90% of Claude-linked output going to GitHub repos w <2 stars
Has always been true. 90% of effort is going to repost with less than 2 stars hasn't changed
SebastianSosa1··on Personal Computer by Perplexity
Comparing to pre-chatgpt workflows makes for useless statements

I want to know pre-"personal computer by perplexity"

SebastianSosa1··on Frontier AI agents violate ethical constraints 30–50% of time, pressured by KPIs
As humans would and do
SebastianSosa1··on GPT-5.2
Community notes on X seems to be one of the highest profile recent experiments trying to address this issue
SebastianSosa1··on Neuroscientists track the neural activity underlying an “aha”
If the cerebral cortex is analogous to a neural network but aha moments or moments of insight depend on the structure of the amygdala plus the hippocampus is there an architecture that vaguely represents the structure between a cerebral cortex amygdala and hippocampus
SebastianSosa1··on Claude Skills
Fair enough, didn't know that it was illegal. The accusation was based on the fact that u released the same day as their release (natural to do so) and with high praise. Thx for the reply I retract my accusation.
SebastianSosa1··on Claude Skills
Love your work but Pretty sure you got paid to endorse them
SebastianSosa1··on Claude Skills
Excellent point, put simply building those preferences and lessons would demand a layer of latent memory, personal models, maybe now is a good time to revisit this idea...
SebastianSosa1··on Ask HN: Who is hiring? (August 2025)
VERIS AI

Cracked engineers (who can research)

Cracked researchers (who can engineer)

Raised 8.6M seed

Building the full vertical for optimizing agents. From scenarios, to simulation, to fine tuning.

Apply to whatever here, you'll likely be wearing all hats anyways: https://jobs.gusto.com/boards/veris-technologies-inc-dadc87f...

Learn more here https://veris.ai/

SebastianSosa1··on Ask HN: What are you working on? (April 2025)
Self Improving Agents with Test Time Reinforcement Learning

https://github.com/CakeCrusher/self_improving_agents