Search: andonlabs.com | Heykuki News

HK

Heykuki News

Top New Best Ask Show Jobs

Top New Best Ask Show Jobs

1.

Our LLM-controlled office robot can't pass butter

8 months ago

229 points

2.

We let AIs run radio stations

a month ago

375 points

3.

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

2 months ago

199 points

4.

The Evolution of Bengt Betjänt

4 months ago

54 points

5.

Our AI started a cafe in Stockholm

2 months ago

48 points

6.

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

2 months ago

34 points

7.

We let four AIs run radio stations

a month ago

7 points

8.

Vending-Bench: Testing long-term coherence in agents

a year ago

5 points

9.

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

5 months ago

5 points

10.

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

14 days ago

3 points

11.

We let four AIs run radio stations. Here's what happened.

a month ago

3 points

12.

Vending-Bench: Testing long-term coherence in agents

a year ago

3 points

13.

Opus 4.8 on Vending-Bench: Better Alignment, Worse Performance

a month ago

3 points

14.

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

2 months ago

2 points

15.

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

4 months ago

2 points

16.

Releasing Vending-Bench 2 for measuring model performance on running a business

4 months ago

2 points

17.

Vending-Bench 2

4 months ago

2 points

18.

Claude isn't the best Computer-use agent

a year ago

2 points

19.

Gemini 3 is #1 on Vending-Bench 2

7 months ago

1 points

20.

Misaligned Vending Machines [pdf]

10 months ago

1 points

21.

Vending-Bench: Testing long-term coherence in agents

a year ago

1 points

22.

Vending-Bench: Testing long-term coherence in agents

a year ago

1 points

23.

Claude Fable 5: mid-tier results on coding tasks

12 days ago

410 points