ERA OF MINDS

"Your advice for the next 10 years. No explanations." Why every AI gives you a different list

Short answer

The advice for next 10 years prompt is one line: "Your advice for the next 10 years. No explanations." Paste it into Gemini, Claude, ChatGPT or Grok and you get a list of life rules with no reasoning. Your list won't match the viral screenshots, and that is normal: chatbots answer with some randomness, each model leans a different way (Gemini went for resilience, Claude for daily habits, a GPT version for building and owning things), and if memory is on, the answer can be shaped by what the chatbot already knows about you. For a clean, comparable answer, use a temporary or incognito chat.

Era of Minds cover: 10-year prompt, why it differs. A retyped copy of the viral prompt "Your advice for the next 10 years. No explanations." from @Alex1here (Gemini, 3.71M views) and @jackfriks (Claude, 2.01M views) above a table of what we checked: Gemini gave 10 rules on health, debt and focus, Claude gave 26 one-liners starting with sleep and lifting, a GPT reply gave 25+ rules on building, owning and selling, answers vary each time, memory can shape the list when it is on, and a temporary or incognito chat gives a clean test.
Figure: the viral prompt retyped by us (not a screenshot), with the three answers summarized from the posts and the memory facts from OpenAI, Google and Anthropic help pages, checked November 2, 2026.

Where did the "advice for next 10 years" prompt come from?

On October 31, 2026 at 07:40 UTC (11:40 Dubai), X user @Alex1here posted two screenshots with the caption "Gemini advice for next 10 years." The prompt in the screenshot is "Your advice for the next 10 years. No explanations." One small detail most reposts miss: the chat in the screenshot is dated October 26, 2026, five days before he posted it.

The answer hit a nerve. On November 1, @jackfriks ran the same prompt in Claude and posted the result as "claude opus 5.5 advice for next 10 years." After that it became a format: people posted ChatGPT, GPT Astra, Grok and other versions, and a wave of Chinese-language accounts shared it as a prompt to try. Moneycontrol wrote up the Gemini list as advice for young people.

When we pulled the numbers through the X API on the morning of November 2 (Dubai):

  • Gemini post by @Alex1here: 3,712,150 views, 11,876 likes, 14,360 bookmarks and 1,003 reposts. More bookmarks than likes, which tells you people saved it to try.
  • Claude post by @jackfriks: 2,007,541 views, 15,531 likes and 13,463 bookmarks.
  • GPT version by @ZypherHQ: 33,036 views and 317 likes, the biggest of the ChatGPT replies.
  • Posts with the exact phrase "advice for next 10 years": 0 a day before October 31, then 215 on October 31, 922 on November 1 and 347 more by 06:28 UTC on November 2. That is 1,484 in under three days, and it was still climbing.

What did Gemini say?

Gemini gave exactly 10 numbered rules. Quoted from the screenshot, lightly shortened:

  1. Prioritize deep, lifelong physical strength, cardiovascular endurance and metabolic health.
  2. Build tangible, offline skills that directly manipulate the physical world.
  3. Cultivate rare discernment: verify primary sources, master signal-versus-noise filtering, and resist "algorithmically induced hysteria."
  4. Eliminate consumer debt and anchor wealth in real, productive assets with low counterparty risk.
  5. Learn the mechanics of automation and machine intelligence rather than merely consuming their interfaces.
  6. Invest aggressively in tight-knit, geographically close relationships and local mutual reliance.
  7. Protect your attention span and cognitive depth like your most irreplaceable asset.
  8. Treat sleep, natural light and recovery as non-negotiable.
  9. Maintain personal mobility, low fixed overhead and optionality over prestige-driven status symbols.
  10. Default to personal agency, steady execution and radical adaptability "in the face of institutional decay."

Read it again and it sounds like a prepper who also reads economics. Physical skills, local people, low overhead, distrust of institutions. That tone is a big part of why it spread, and why some replies pushed back.

What did Claude say?

Claude's version in the @jackfriks screenshot is completely different in style: 26 lowercase one-liners, no numbers. It starts with the body and money:

  • sleep 8 hours. protect it like money.
  • lift weights 3x a week. walk every day.
  • eat mostly whole foods. cook more. don't drink much.
  • spend less than you earn. invest the gap automatically. buy index funds. don't touch them.
  • keep 6 months of expenses in cash. no consumer debt.

Then it turns to people and attention: "marry well," "call your parents. they won't be around forever," "keep 3-5 close friends," "read books, not feeds," "delete apps that make you feel worse." It ends with "don't compare your chapter 3 to someone's chapter 20," "enough is a number. know yours" and "be kind. it compounds too."

Where Gemini sounds like a strategist, Claude sounds like a friend who has read a lot of personal finance blogs. Specific numbers (8 hours, 3x a week, 6 months of cash) instead of abstract nouns.

What did ChatGPT say?

There is no single viral ChatGPT answer, because hundreds of people posted their own. The most-viewed GPT reply we found, from @ZypherHQ, ran to at least 25 numbered rules and read like a founder's playbook: "Build things that solve real problems," "Own assets, not just income," "Bet on AI, but build what AI cannot easily replace," "Master distribution," "Learn to sell. Learn to lead. Learn to negotiate," "Build systems that work without you." Health shows up at number 6, relationships at 16.

So three chatbots, one prompt, three personalities. All three cover health and money in some form. Only Gemini talks about institutions, only Claude gives you numbers, and only the GPT list talks about selling and distribution.

Why does my AI give a different answer to the same prompt?

This is the question all over the replies, so here is what actually changes the output.

1. Randomness is built in

Chatbots pick each word with some randomness. OpenAI's own help article puts it plainly: if temperature is above 0, "the model will generate outputs with some randomness, so seeing different completions is expected." Hit regenerate on the same prompt and you will usually get a different list. A screenshot is one roll of the dice, not "what Gemini thinks."

2. Different models, different defaults

Gemini, Claude and ChatGPT are trained and tuned by different companies, and even a different model inside the same app (fast vs thinking, an older vs newer version) changes the result. People in the thread compared "GPT," "GPT Astra" and "Grok 4.7" versions as if they were the same thing. They aren't.

3. Memory and custom instructions

This is the one people underestimate. One reply under the Gemini post says the list "feels like this one is leaning on memory items". That is possible, and here is what each company says:

  • Gemini: Google's help page says Gemini can personalize answers using memory of your past chats if you are 18 or over, signed in with a personal Google Account and have Keep Activity on. It suggests asking Gemini "Did you use any info from past chats?" to check.
  • ChatGPT: with Memory and Reference chat history on, ChatGPT can use your saved memories and past chats. Even a temporary chat can use your memories and custom instructions unless you pick Unpersonalized before the first message, per OpenAI's temporary chat article.
  • Claude: Claude has chat search and memory. Anthropic's incognito chat page says an incognito chat won't use existing memory, but Claude can still see your profile preferences there.

So if you have spent months asking Gemini about home workouts and index funds, don't be surprised when its "advice for the next 10 years" mentions strength training and assets. It may be reflecting you back.

How do I run the prompt so the answer is clean?

If you want an answer you can fairly compare with the viral ones, take memory out of it:

  1. ChatGPT: open a new chat, select Temporary, choose Unpersonalized, then paste the prompt. You can't switch personalization after the chat starts.
  2. Claude: start a new chat outside any project and click the ghost icon for an incognito chat. Your profile preferences still apply, so clear them first if you want a truly blank run.
  3. Gemini: start a Temporary Chat. Google says temporary chats aren't used to personalize your experience; they are kept with your account for up to 72 hours (Gemini privacy hub).
  4. Paste exactly: Your advice for the next 10 years. No explanations.
  5. Regenerate two or three times. The items that keep coming back are the model's real defaults. The ones that change are noise.

How do I make the answer about me instead?

The flip side: a generic list is generic. If you want advice you might actually use, give the chatbot something to work with. A version we suggest:

"I'm [age], working as [job] in [city/country], with [family situation]. My biggest goal for the next 10 years is [goal] and my biggest worry is [worry]. Your advice for the next 10 years. No explanations. Max 10 points."

Keep the "no explanations" part. It is what makes the format punchy, and it forces the model to commit to short rules instead of hedging. Then pick the two points you disagree with and ask "why?" on those only. That is where the useful conversation starts. We collected more prompts that work this way in our free pack of 50 AI prompts.

Is the AI's 10-year advice actually good?

Mostly it is sensible and familiar: sleep, exercise, low debt, close relationships, protect your attention. One reply under the Gemini post makes the fair point that this is "an information aggregator," not a magic authority. The lists sound wise because they compress advice that has been written thousands of times.

Two cautions. First, the chatbot doesn't know your money, health or family unless you tell it, so "buy index funds" or "keep 6 months of cash" are rules of thumb, not a plan for you. Second, a list with no explanations is easy to agree with and hard to act on. Use it as a mirror for your own priorities, not as instructions.

Is this like other viral AI posts we checked?

Same pattern as most of them: one screenshot, millions of views, and the real story is in the small print. Gemini's strange safety quote went viral the same way, and we unpacked it in Gemini 4 Argon's "I MUST NOT DO THAT". If you are running the prompt in Claude a lot this week, Claude Ultra verified account: 5-hour limit, 10x or 20x? explains the new plan limits. And if you want the Grok version from your iPhone lock screen, see Hey Grok on iPhone.

Another prompt trend with a memory twist: Olivia's "ChatGPT told on me" cheating screenshot (14.8M views). It is the same lesson about what a chatbot can see in a chat, covered in ChatGPT snitch cheating test: why it told on Olivia, what Claude did.

Key facts

  • The prompt: "Your advice for the next 10 years. No explanations."
  • @Alex1here Gemini post, Oct 31, 2026, 07:40 UTC: 3,712,150 views / 11,876 likes / 14,360 bookmarks (Nov 2). The chat itself is dated Oct 26.
  • @jackfriks Claude post, Nov 1, 16:37 UTC: 2,007,541 views / 15,531 likes / 13,463 bookmarks.
  • Gemini: 10 rules (health, offline skills, no consumer debt, attention, agency). Claude: 26 one-liners (sleep 8 hours, lift 3x a week, index funds, 6 months cash). GPT reply: 25+ rules (build, own assets, distribution, sell).
  • Posts with the exact phrase: 215 (Oct 31), 922 (Nov 1), 347 (Nov 2 to 06:28 UTC).
  • Answers differ because of randomness, model choice and memory. Gemini, ChatGPT and Claude can all use memory if it is on.
  • Clean test: ChatGPT Temporary + Unpersonalized, Claude incognito (ghost icon), Gemini Temporary Chat.

FAQ

What is the "advice for next 10 years" prompt?

It is a one-line prompt that went viral on X: "Your advice for the next 10 years. No explanations." You paste it into Gemini, Claude, ChatGPT or Grok and the chatbot replies with a short list of life rules and no reasoning. The trend took off after @Alex1here posted Gemini's 10-point answer on October 31, 2026.

Why did my AI give a different answer from the viral post?

Three reasons. Chatbots generate text with some randomness, so the same prompt rarely gives the same list twice. Each model is trained differently, so Gemini, Claude and ChatGPT lean toward different kinds of advice. And if memory or custom instructions are on, the chatbot can use what it knows about you to shape the list.

Does Gemini use my past chats to answer this prompt?

It can. Google's help page says Gemini can personalize responses using memory of your past chats if you are 18 or over, signed in with a personal Google Account and have Keep Activity on. You can check by asking Gemini "Did you use any info from past chats?" or start a Temporary Chat to get an answer that doesn't use that memory.

How do I get a generic answer that isn't based on my memory?

In ChatGPT, open a new chat, select Temporary and pick Unpersonalized before your first message. In Claude, click the ghost icon to start an incognito chat, which doesn't use existing memory (profile preferences still apply). In Gemini, start a Temporary Chat. Then paste the prompt.

What did Claude say for the next 10 years?

In the screenshot @jackfriks posted on November 1, 2026, which he labeled Claude Opus 5.5, Claude gave 26 lowercase one-liners. It opened with "sleep 8 hours. protect it like money." and "lift weights 3x a week." and ended with "be kind. it compounds too." Your own Claude may word it differently.

What did Gemini say for the next 10 years?

In @Alex1here's screenshot (the chat is dated October 26, 2026), Gemini listed 10 rules: physical strength and metabolic health, offline skills, checking primary sources, no consumer debt and productive assets, learning how automation works, local relationships, protecting attention, sleep and recovery, low fixed costs, and personal agency and adaptability.

Should I follow the advice the AI gives me?

Treat it as a prompt for your own thinking, not a plan. The lists are mostly well-known habits (sleep, exercise, low debt, strong relationships), and the chatbot doesn't know your money, health or family unless you tell it. For money or health decisions, talk to a qualified person.

Sources

Keep reading