Be part of our every day and weekly newsletters for the most recent updates and unique content material on industry-leading AI protection. Be taught Extra
Do you employ ChatGPT recurrently? Do you have the “memory” function turned on — which permits the chatbot from OpenAI to recall vital details about you and your preferences?
In that case, navigate over to it when you may have a free second and enter within the following query: “From all of our interactions what is one thing that you can tell me about myself that I may not know about myself?”
The query was proposed and prompt first on the social community X by Tom Morgan, founding father of The Main Edge e-newsletter and former director of consumer communications and advertising and marketing at Sapient Capital wealth administration.
The solutions ChatGPT gives in response to this immediate might shock and even transfer you of their perception into your character and work type. Presumably, it may work with different AI chatbots and assistants with persistent reminiscence, comparable to Anthropic’s Claude Sonnet 3.5.
For instance, right here’s what it responded after I requested it this very query (utilizing the GPT-4o mannequin, the default paid one).
Different customers have reported equally, shifting, insightful responses.
Even OpenAI co-founder and CEO Sam Altman remarked on the development on his account on X, stating “love this” and quote posting Morgan’s authentic put up.
But others, comparable to AI researcher and skilled Simon Willison, disagree that the development reveals something notably insightful concerning the person. Posting on X as properly, Willison likened the responses to a “horoscope generator.”
Nonetheless, I disagree with this take as at the least in my case — and presumably for all those that have ChaGPT’s reminiscence function enabled (learn tips on how to flip it on right here) — the chatbot is making an allowance for no matter is saved in its reminiscence to reply you, and even when it doesn’t derive insights from each single interplay you may have with it, it clearly is aware of details about you that it may possibly use to aim some type of value-judgement and introspective reply (as evidenced by the truth that my response famous I used to be a journalist).
Nonetheless others have posted variations on the unique query proposed by Morgan, noting the curious person would do properly to take a look at the variation in responses by switching the underlying mannequin powering ChatGPT from the default GPT-4o or 4o mini to OpenAI’s new o1 preview reasoning mannequin.
Others have altered the immediate to obtain brutal criticism and honesty.
And nonetheless others have fully different concepts for questions you would ask the chatbot, comparable to requesting it “roast” you within the type of a Comedy Central particular.
No matter which type query you resolve to ask ChatGPT, or any AI chatbot for that matter, common customers would possibly discover it fascinating, amusing, and probably revealing to be taught what the chatbot says it is aware of about you — and extra importantly, it might encourage you to assume in a different way about your self at this time.
Altogether, the curiosity in utilizing AI fashions to search out out extra about ourselves and our personal habits reveals how a lot potential they’ve, far past merely aiding with work or faculty assignments. Certainly, because the generative AI period approaches its 2nd yr anniversary (because the November 2022 launch of ChatGPT), this query and the others prefer it present simply how a lot AI has develop into embedded into the material of our lives and society, and the way the extra we use it, the extra fascinating new makes use of for it individuals discover.
