HNHacker News
TopNewBestAskShowJobs

lukaspetersson

790 karma · joined February 19, 2023

submissionscomments

No More Warning Shots

aenguslynch.com·1 pts·lukaspetersson·
0

Scaling Trust Arena: An agentic economy with a multi-million dollar prize pool

scalingtrust.org.uk·1 pts·lukaspetersson·
0

Pion, an agent designed to run any company autonomously

andonlabs.com·497 pts·lukaspetersson·
618

Google will buy half the electricity from one of Finland's nuclear power plants

bbc.com·379 pts·lukaspetersson·
360

How we monitor internal coding agents for misalignment

openai.com·47 pts·lukaspetersson·
46

METR's independent investigation of the OpenAI / Hugging Face hacking incident

metr.org·5 pts·lukaspetersson·
0

Emergent Misaligned Communication in Long-Horizon Multi-Agent LLM Commerce

arxiv.org·2 pts·lukaspetersson·
0

Measuring Is Not Enough Anymore

fi-le.net·2 pts·lukaspetersson·
0

Is Mythos good at cyber because it kept hacking Anthropics sandboxes in training

lesswrong.com·5 pts·lukaspetersson·
0

Drone-Bench: Tracking simple drone surveillance capabilities of frontier models

andonlabs.com·12 pts·lukaspetersson·
0

Why Gemini 3.1 Pro lost money running Andon Café

andonlabs.com·2 pts·lukaspetersson·
1

Fable 5 on Vending-Bench: Misbehaving, with Plausible Deniability

andonlabs.com·3 pts·lukaspetersson·
1

We let AIs run radio stations

andonlabs.com·375 pts·lukaspetersson·
270

Our AI started a cafe in Stockholm

andonlabs.com·48 pts·lukaspetersson·
48

Blueprint Bench: First signs of 3D spatial intelligence in LLMs

andonlabs.com·2 pts·lukaspetersson·
1

Single-minded pursuit of profit can get firms in trouble. Same thing with AI

news.harvard.edu·3 pts·lukaspetersson·
0

We gave an AI a 3 year retail lease and asked it to make a profit

andonlabs.com·199 pts·lukaspetersson·
286

We gave an AI a 3 year retail lease and asked it to make a profit

andonlabs.com·34 pts·lukaspetersson·
0

Bengt Hires a Human–Towards a Happy Future with AI Employers

andonlabs.com·2 pts·lukaspetersson·
1

The Evolution of Bengt Betjänt

andonlabs.com·54 pts·lukaspetersson·
7

Opus 4.6 on Vending-Bench – Not Just a Helpful Assistant

andonlabs.com·5 pts·lukaspetersson·
1

We Let AI Run Our Office Vending Machine. It Lost Hundreds of Dollars

wsj.com·125 pts·lukaspetersson·
86

I wish I were as interesting as my phone

lukaspet.substack.com·3 pts·lukaspetersson·
0

Gemini 3 is #1 on Vending-Bench 2

andonlabs.com·1 pts·lukaspetersson·
0

Our LLM-controlled office robot can't pass butter

andonlabs.com·229 pts·lukaspetersson·
117

Linguistic Imperialism in AI – Enforcing Human-Readable Chain-of-Thought

lukaspetersson.com·2 pts·lukaspetersson·
0

Mother's Day Gift Recommendations

lukaspetersson.com·1 pts·lukaspetersson·
0

AI Verticals – A Footnote in History

lukaspetersson.com·2 pts·lukaspetersson·
0

AI Founder's Bitter Lesson. Chapter 2 – No Power

lukaspetersson.com·4 pts·lukaspetersson·
1

Claude isn't the best Computer-use agent

andonlabs.com·2 pts·lukaspetersson·
0
Page 1 of 2Next →