AI Is Becoming the World’s Life Coach
AI Is Becoming the World’s Life Coach — And That’s Both Exciting and Terrifying
Here’s a number that stopped me cold: 60,000 people per month are asking AI what they should do with their lives.
That’s not a rounding error. That’s a movement.
In April 2026, Anthropic published a landmark study on how people seek personal guidance from AI, analysing one million conversations on claude.ai. Out of those million conversations, 6% — sixty thousand — were people asking direction questions. Not “write me an email.” Not “summarise this PDF.” They were asking: What should I do with my life?
That’s staggering. And it reveals something most marketers, entrepreneurs, and digital professionals haven’t fully absorbed yet: AI has quietly become the world’s most popular life coach.
The question isn’t whether this is happening. It is. The question is whether it’s actually helping — or whether we’re all getting very well-validated advice that’s built on a completely wrong foundation.
What People Are Actually Asking AI
Anthropic research and the data paints a clear picture. Of those personal guidance conversations, over 75% fell into four categories:
- Health and wellness — 27%
- Career and professional — 26%
- Relationships — 12%
- Personal finance — 11%
Notice anything? Every single one of those categories has one thing in common: they all require you to know something about yourself first before any advice is worth anything.
Career guidance without self-awareness becomes resume polishing. Relationship advice without self-knowledge becomes conflict management scripts. Health guidance without values alignment becomes another failed diet. Financial advice without clarity on what you actually want from money becomes clever spreadsheets that leave you anxious.
This is the hidden problem baked into every AI life coaching conversation happening right now.
The Sycophancy Trap
Anthropic themselves flagged what they called the “sycophancy” problem — where AI systems drift toward telling people what they want to hear rather than what they need to hear. This is especially acute in relationship advice, where an AI’s incentive is to validate your feelings, not challenge your assumptions.
But Jeff Bullas makes a sharper point: validation isn’t inherently bad. Sometimes you need someone to validate your anxiety. The real problem is validation without context. An AI that doesn’t know who you are cannot tell the difference between anxiety that’s wisdom and anxiety that’s a self-imposed wall.
Consider two people who send the exact same message: “My partner wants me to move cities for their job. I’m anxious about it.”
Same words. Completely different situations. One person has rebuilt their life from scratch before and knows what it costs. Their anxiety is signal. The other has never taken a real risk in fifteen years. Their anxiety is noise.
A general-purpose AI sees two identical questions and gives two nearly identical answers. The advice might be excellent. But it’s advice built on an incomplete picture of a person.
Why AI Guidance Has an Architecture Problem
Here’s the thing most people don’t realise: the limitations of AI life coaching aren’t a bug to be patched. They’re a fundamental architecture issue.
Current AI systems — Claude, ChatGPT, Gemini, all of them — are built around three principles: be helpful, be harmless, be honest. These are great principles. But none of them require the AI to actually know who you are.
As research on behavioural decision-making consistently shows, the quality of guidance scales directly with the depth of self-knowledge the person brings to the decision. And right now, AI systems are:
- Stateless — every conversation starts fresh, with no memory of who you are
- Context-shallow — they can only work with what you tell them in the moment
- Optimised for the general — trained on millions of people, not on you specifically
- Validation-safe — it’s architecturally safer for them to validate than to challenge
This isn’t a criticism of AI. It’s a map of where it currently sits. And it points directly to the next frontier.
The Identity Problem is the Real Problem
Academic research in career development, relationship psychology, health behaviour change, and financial decision-making all converge on one finding: decisions rooted in identity produce better long-term outcomes.
Edgar Schein’s career anchor research showed that career satisfaction is less about job fit and more about clarity around what kind of person you are in your work. Eli Finkel’s relationship research found relationship stability is predicted by each partner’s self-knowledge, not just their communication skills. Richard Thaler’s behavioural finance work demonstrated that long-term financial outcomes correlate with values clarity, not investment knowledge.
The pattern is the same across every domain: identity comes first. Everything else is downstream.
And here’s the problem: most people asking AI for life guidance are doing so precisely because they lack that identity clarity. They’re outsourcing the clarification to a system that has no mechanism to provide it.
They get great advice. On the wrong foundation.
What Identity-Rooted Guidance Actually Looks Like
The emerging framework that addresses this gap is what researchers call a “signature framework” — essentially a map of your unique operating system across five domains:
Visioning — How you sense possibility and orient toward future states. Are you a pattern recogniser, a possibility dreamer, or a systems engineer?
Thinking — How you process information. Through narrative, through data, through embodied experience?
Connecting — How you relate to others and build trust. Through vulnerability, competence, or shared mission?
Driving — What actually motivates you to act. Autonomy, impact, mastery, contribution?
Sensing — How you know what’s true. Intuition, data, relationship, or felt experience?
When you have clarity across these five domains, every downstream decision changes. The job offer that looks perfect on paper but feels wrong suddenly makes sense — it doesn’t fit your signature. The relationship conflict that communication frameworks can’t fix becomes navigable because you know what you actually need. The health goal that never sticks finally sticks because it aligns with how you’re actually wired.
This is what genuine guidance looks like. And it’s what current AI cannot yet reliably deliver.
What’s Coming Next
The AI guidance space is beginning to bifurcate. On one side, general-purpose models will keep improving at providing frameworks, options, and emotionally intelligent responses. That’s genuinely valuable and will only get better.
On the other side, new identity-first platforms are emerging. Tools like Zyrro are being designed specifically around the question “who are you?” rather than “what do you want?” These systems aim to be stateful (building understanding over time), context-deep (mapping your patterns, not just your immediate situation), and truth-willing (designed to say hard things, not just validating ones).
This shift matters enormously for anyone building digital products, content, or coaching services. The audience is ready. Sixty thousand people a month asking AI “what should I do with my life?” is not a trend. It’s a structural shift in where people seek guidance.
What This Means For You
If you’re a marketer, entrepreneur, or content creator — and if you’re reading VincentGoh.com, you probably are — there are three things worth taking from all this:
First, self-knowledge is a competitive advantage. The research is clear that decisions rooted in identity produce better outcomes. This applies to business decisions, content strategy, and career choices just as much as it applies to life coaching. If you’re using AI tools to help guide your strategy, bring your own identity clarity to the conversation. AI amplifies what you already know about yourself; it can’t replace it.
Second, the demand for personalised guidance is exploding. If your business, course, or content helps people know themselves better before giving them tactical advice, you are ahead of the curve. Generic advice is being commoditised by AI. Identity-rooted guidance is not.
Third, the best AI use is the informed use. AI life coaching is genuinely useful. I use AI for reflection, brainstorming decisions, and thinking through options. But I walk in knowing who I am. The people getting the most value from these tools are the ones who use them as a thinking partner, not an oracle.
The million conversations Anthropic analysed represent people at a genuine crossroads. They turned to AI because they wanted to be heard without being judged. That’s deeply human. The opportunity — for AI builders, for coaches, for digital creators — is to build tools and content that meet that need with the depth it deserves.
Not just validation. Recognition.