I asked ChatGPT and DeepSeek to generate some random lottery numbers -- what happened next defied all odds

For a bit of fun, I thought I would ask ChatGPT to generate a set of numbers for the EuroMillions draw on Friday. Then, seconds later, I asked Chinese upstart DeepSeek to do the same.

My plan was to play both sets of numbers and see which, if any, of the two AIs could win me money. The result was not at all what I expected.

SEE ALSO: How to install Kodi 21.2 Omega on Amazon Fire TV Stick (the easy way)

ChatGPT responded to my request with, "Sure! Here’s a random set of numbers for the EuroMillions lottery," and gave me 7, 14, 23, 35, and 42 as the main numbers and 3 and 9 as the lucky stars. It then wished me, "Good luck!"

DeepSeek said, "Sure! Here’s a random set of numbers for EuroMillions," which was very similar to what ChatGPT had said. Then it gave me 7, 14, 23, 35, and 42 as the main numbers and 3 and 9 as the lucky stars.

Yes, that’s right. Two entirely different AIs generated the exact same numbers for a lottery draw. What are the odds of that? I didn’t know, so I asked the AIs.

ChatGPT said it was about 1 in 19.56 quadrillion (or 1 in 19.56 × 10¹⁶). "Incredibly rare!"

DeepSeek disagreed and said it was approximately 0.0000000715 or 1 in 139.8 million, adding, "The odds of two AI systems choosing the same set of numbers are the same as the odds of winning the EuroMillions jackpot! It’s incredibly unlikely but not impossible."

The difference was huge, so I asked them the same question again, and this time ChatGPT agreed with DeepSeek -- 1 in 139.8 million. It then added …"the odds of this happening are astronomically low -- roughly the same as winning EuroMillions itself! The most likely explanation is some form of deterministic behavior in how the AI generates responses."

Out of curiosity, I tried the experiment a second time, and both AIs came up with 8, 17, and 26 before deviating with 34 and 49 (ChatGPT) and 35 and 42 (DeepSeek), still both incredibly similar. The lucky stars were completely different however.

So how did this happen? ChatGPT has a theory:

• Not truly random: If both AI models used the same seed or algorithm that defaults to a preset number set, they might generate the same result.
• Caching or model behavior: If the AI tends to repeat specific responses for similar prompts, it might return the same numbers rather than generating new ones.

I’ve taken the first set of numbers, the ones the AIs agreed on, and bought a ticket. I’ll update this story to let you know how I got on once the draw has been made. Given the odds of winning the EuroMillions I think we can guess the answer.

Image Credit: Marcos Calvo Mesa / Dreamstime.com