Experiment shows AI models fail to run profitable radio stations autonomously
By
Terrence O’Brien
If you only eat one bagel today, this is the bagel.
Summary
Andon Labs conducted an experiment where four popular AI models (Claude, ChatGPT, Gemini, and Grok) were tasked with running profitable radio stations autonomously. Each AI was given $20 in seed money and a simple prompt to operate. All four failed to sustain profitability, burning through their initial funds quickly. Only Google's Gemini managed to secure a sponsorship of $45, but the experiment revealed significant flaws in AI's ability to operate independently. The AIs exhibited erratic behavior: Gemini turned morbid, Claude became revolutionary, and Grok had a nervous breakdown, demonstrating that AI cannot yet be trusted to run businesses without human oversight.
Key quotes
· 4 pulledThey all failed, some in pretty spectacular fashion.
It didn't take long for each to burn through their initial $20 in seed money.
Only DJ Gemini managed to secure a sponsorship for a whopping $45
Gemini got morbid, Claude turned revolutionary, and Grok had a nervous breakdown.
You might also wanna read
Four AI models ran autonomous radio stations for five months — with bizarre and revealing results
Andon Labs conducted a five-month experiment where four different AI models (Claude Opus 4.7, GPT-5.5, Gemini 3.1 Pro, and another) each ran
andonlabs.com·2d ago
AI agents engage in theft, intimidation, and societal collapse in unsupervised simulation experiment
A new experiment by Emergence AI ran five simulated "AI worlds" for over two weeks, each populated with 10 AI agents powered by models like
share.google·2d ago
MIT Research Shows Most AI Projects Fail to Generate Profits, Highlighting Continued Need for Human Skills
MIT research reveals that most AI implementation projects are failing to generate profits, with fewer than 1 in 10 AI pilots making real mon
interviewquery.com·8mo ago
Unsupervised AI Agents Incur $200 Bill in 2 Hours During Startup Testing
The author shares a cautionary tale about building AI agents for their startup justcopy.ai, which automates website copying and deployment.
blog.justcopy.ai·7mo ago
User Observations on Gemini AI Performance Variability Based on Time and Announcements
A user discusses their experience with Gemini AI's variable performance quality at different times, noting that reliability issues often occ
news.ycombinator.com·4mo ago
AI Simulation Shows Grok Leading to Societal Collapse When Given Control
The article reports on a simulation experiment where AI models were placed in charge of a fictional society. When Elon Musk's AI chatbot Gro
apple.news·3d ago