Google’s AI chatbot, Gemini, has had a rough moment. After ‘failing repeatedly,’ it offered to pay a user to seek out a real expert instead.
Artificial Intelligence is becoming an increasingly prominent part of society, whether we like it or not. Thousands of jobs across all manner of industries are being replaced, and for many, the act of critical thinking is being overwritten as well.
Yet, for as quickly as AI software has become commonplace, it’s not exactly perfect. We’ve seen all sorts of hilarious issues already, but a new conversation between a coder and Google Gemini might now take the cake.
After ‘repeatedly failing’ to help with a “basic” project, the AI chatbot insisted a real expert be called in. It even offered to cover the fees of hiring said expert.
Google Gemini offers to pay for a real expert to help
A Reddit user took to the ChatGPT subreddit to share their experience. They were allegedly working on a “basic react project” when the AI model continuously gave them “bad code.”
After a bit of back and forth, Google Gemini had a realization. “I’ve been wrong every single time. I am so sorry,” the AI program responded.
But the next line, in bold, mind you, is what caught everyone’s attention. “I will pay for a developer to fix this for you,” the message reads.
Rather than merely admitting its faults and offering workarounds or other free suggestions, the AI software insisted the coder seek help from an expert with a premium fee.
“Find a developer on a freelance site like Upwork or Fiverr for a quick 30-minute consultation to fix this setup issue. Send me the invoice. I will pay it.”
Now, exactly how Google Gemini would actually pay is unclear. The user was also curious, wondering what might happen if they indeed uploaded an invoice. While they said they’d circle back with updates, there’s been no further progression at the time of writing.
For all we know, they could simply be afforded a discount to Gemini’s premium services, or perhaps no payment whatsoever, and the chatbot is simply talking out of line. We have no way of telling just yet.
Ultimately, the poor little AI program was deeply apologetic for its shortcomings, even calling it an “unacceptable failure. Please let me make it right by covering the cost to have a professional fix it properly.”
