I was wrong about ChatGPT.
Six months ago, when the hype first hit, I did what most analysts do—I panicked quietly and told myself I'd figure it out later. I saw the headlines: "AI will replace half the jobs." I watched colleagues in finance start using it to draft reports. And I thought: this is either the biggest threat to my career or the smartest tool I haven't learned yet.
I chose to ignore it for a while. Classic move.
Then one Tuesday morning, stuck in the Kalyan-to-Mumbai train (delayed by 40 minutes, as usual), I opened ChatGPT on my phone out of boredom. I asked it to explain a complex financial concept I was struggling to communicate to a non-technical stakeholder. The response wasn't perfect, but it was clear. It gave me a starting point. And something shifted.
I've spent the last six months actually using this tool, not just reading about it. And what I've learned isn't what the headlines promised. It's messier, more human, and honestly? A lot more interesting.
What I Got Wrong About AI and Work
Here's what I believed three months ago: ChatGPT would make certain jobs obsolete. Spreadsheet jockeys like me would be out. Content writers would be replaced. Customer service would be fully automated. The future was a race to retrain or get left behind.
This narrative felt true because it's scary. And fear sells. But here's the thing—it's also lazy thinking.
The "Replacement" Myth
I spent hours reading about how AI would eliminate jobs the way ATMs reduced bank tellers. The logic is clean: if a machine can do X, humans don't need to do X anymore. Seems obvious, right?
Except it's not how work actually happens.
A few weeks in, I realized that ChatGPT didn't replace my analysis—it replaced the boring parts of my analysis. Drafting the first version of a report structure. Explaining methodology in simple terms. Finding examples for a concept. Writing the "here's what this means" summary that takes an hour and feels pointless.
What it didn't do: ask the right questions. Understand the context of why we're analyzing something. Catch the logical flaw in my argument. Push back and say, "Are you sure about this conclusion?"
That's still me. That's still the human job.
And honestly? Removing the tedious parts didn't devalue me. It actually revealed what I'm actually paid for. Decision-making. Judgment. Context. The stuff that requires you to know the industry, know the stakeholders, know what happened last quarter and why it matters.
The "Everyone Must Retrain" Panic
I also believed—deeply—that if you didn't immediately become an "AI expert," you'd be obsolete. So I got defensive. I started reading every Medium article about prompt engineering. I thought about taking a course. I felt the pressure to become someone who "knows how to use AI" the way people suddenly needed to "know Excel" in 2005.
This is where I was truly, embarrassingly wrong.
The people I actually respect and work with? They're not taking AI courses. They're just using it. Like they use email. Like they use Google. They ask it questions, they check the answers, they iterate. Sometimes it saves them 2 hours. Sometimes it's useless and they move on.
The pressure to become an "AI expert" is mostly just marketing. What actually matters is understanding what the tool can and can't do, then using it for the jobs where it genuinely helps. That's not a course. That's just... trying it out.
| What I Thought ChatGPT Would Do | What It Actually Does | Impact on My Work |
|---|---|---|
| Replace me as an analyst | Handle template writing, initial structure, explanations | Saves 2–3 hours on writing, but still needs my judgment on conclusions |
| Make me instantly obsolete without training | Require regular use and critical thinking to integrate well | My domain expertise is now more valuable, not less |
| Solve complex problems on its own | Help me think through problems faster, or generate options I refine | I'm faster at ideation and drafting, but still responsible for accuracy |
| Be correct most of the time | Be confident and wrong sometimes, especially on facts and numbers | I need to verify everything, so it's a thinking partner, not a fact-checker |
What Actually Happened at Work
Let me get specific about how this tool actually changed my day-to-day work at Morningstar, because the real story is less dramatic and more useful than the hype suggests.
1. I Got Faster at Writing (But Not Smarter)
Our reports used to take a full day to write from scratch. Research, writing, revising, sending to stakeholders for feedback. Now?
I spend the morning researching and analyzing. That part hasn't changed. Then, instead of staring at a blank screen for an hour, I give ChatGPT my findings and ask it to draft the "Overview" section. It takes 10 minutes. The output is roughly 60% usable. I rewrite about 40% of it to match our tone and add the specific context the AI couldn't know.
Net result: I save maybe 2-3 hours per report.
This is genuinely good. But here's what's not happening: I'm not deeper in my thinking. I'm not understanding the data better. I'm not making smarter decisions. I'm just writing faster.
The gap between "faster" and "better" matters way more than the hype acknowledges.
2. I Stopped Making Dumb Mistakes Faster
This one surprised me.
Before ChatGPT, when I'd finish a report, I'd read it once, maybe twice. I'd catch 70% of the errors. The other 30% would get through—awkward phrasing, logical gaps, a sentence that didn't quite make sense.
Now, I paste my draft into ChatGPT and ask it to review for clarity and logic. It catches most of the weird sentences. Sometimes it catches actual logic errors ("You said this fund returned 15% but earlier you said it returned 12%—which is correct?").
Does it catch everything? No. But it's like having a first-pass editor who never gets tired. And that saves me from sending draft-quality work to stakeholders.
The real skill here isn't "knowing ChatGPT." It's knowing when and how to use it as a filter. That's just experience.
3. It Exposed What I Actually Don't Know
This is the uncomfortable one.
Using ChatGPT forced me to get specific about things I thought I understood. When I'd ask it to explain a concept, if the explanation was wrong, I'd notice. If the explanation was right but shallow, I'd realize I couldn't go deeper. If I couldn't verify what it was saying, I'd realize I had a gap in my knowledge.
In a weird way, ChatGPT became a mirror for my own knowledge. And that's useful for career growth—you can't improve what you don't measure.
For example, I thought I understood mutual fund categorization well. Then I asked ChatGPT to explain the difference between "large-cap" and "mega-cap" funds in Indian context (with INR examples), and it gave me a response that sounded right but I couldn't verify. So I had to actually research it. Turns out my mental model was approximately correct but had gaps I didn't know existed.
Would I have found those gaps without the tool? Eventually, maybe. But probably not until I made a mistake in front of someone important.
How Your Career Changes (The Honest Version)
Okay, so what does this actually mean for someone like us—Indians in their 20s-30s, working in knowledge work, worried about job security?
Here's what I genuinely believe now, after six months of real use:
Your Skill Value Is Shifting, Not Disappearing
The parts of your job that are pure execution—writing emails, drafting documents, creating first drafts, formatting reports, explaining concepts—these are becoming cheaper and faster. That's real. But those were probably only 40% of your actual value anyway. The other 60%—judgment, taste, knowing what matters, understanding context, building relationships—that's still entirely human.
The catch: you have to actually develop that 60%.
If you've been coasting because you're fast at emails and reports, ChatGPT is going to feel like a threat. Because it's faster at that than you are. But if you've been using those tasks as a foundation to build deeper thinking and stakeholder relationships, the tool just frees you up to do more of the actually valuable work.
This is why I think it's not "Will AI replace me?" but "Will I replace the boring parts of my job with something better?"
Domain Expertise Just Got More Valuable
Here's the thing nobody talks about: ChatGPT is generalist. It knows a little about everything. But it doesn't know your business, your market, your customers, your history.
When I use ChatGPT at Morningstar, the best use cases are when I combine its speed with my specific knowledge. I know Indian fund categories. I know which funds are overhyped. I know what Morningstar's reputation is and what we can and can't say. ChatGPT helps me articulate that knowledge faster. But without my domain expertise, the output would be generic and wrong.
This means: if you've spent three years learning your field deeply, ChatGPT doesn't make that irrelevant. It makes it more important. Because you're now the filter between the AI-generated output and actual value.
That's more valuable, not less. The market just doesn't realize it yet.
Speed Is the New Baseline
Here's the uncomfortable truth: everyone is going to have access to ChatGPT. Your competitor has it. Your colleague has it. The person applying for your promotion has it.
So "using ChatGPT" isn't a differentiator. It's just the minimum bar.
What differentiates you is: Do you know when to use it? Do you ask better questions? Do you have better judgment about what the output means? Can you combine AI-generated speed with human insight?
In other words: the tool doesn't replace skill—it amplifies it. If you're already thoughtful and careful, ChatGPT makes you faster and more careful. If you're sloppy, it just makes you faster at being sloppy.
| Skill Type | Impact of ChatGPT | What to Focus On |
|---|---|---|
| Execution (writing, formatting, drafting) | AI is faster and cheaper | Use the tool, don't compete on this |
| Domain knowledge (industry-specific expertise) | Becomes more valuable (you're the filter) | Deepen your expertise in your niche |
| Judgment (knowing what matters, context) | Cannot be automated, purely human | Build this through experience and reflection |
| Relationships (trust, communication, influence) | Not affected by AI, but you have more time for it | Invest in relationships, not just tasks |
What I'm Still Figuring Out
I don't have all the answers here. And I'm suspicious of anyone who does.
I'm still learning when ChatGPT is genuinely useful versus when I'm just using it because it's there. I'm still making mistakes with it—taking its confident-sounding wrong answers as fact and having to catch them later. I'm still probably underestimating both how good it will get and how people will adapt to using it.
But here's what I know: The job market doesn't disappear overnight. It shifts. Some roles become less valuable. New ones emerge. People who panic and do nothing get hurt. People who adapt thoughtfully come out ahead.
And the truth is: if you're reading this, you're probably already adapting. You're thinking about it. You're trying to understand what it means for you. That's already half the battle.
My Perspective
I spend about 90 minutes a day on the Mumbai local train. That's when I think about stuff like this—when I'm standing, crushed between 200 other people, staring at my phone or out the window.
Somewhere between Kalyan and Dadar, it hit me: all the panic about AI is from people who are afraid of being replaced. But replacement isn't how work changes. Work changes when tools make certain skills cheaper, so people move to skills that tools can't make cheaper. That's just evolution.
What surprised me most? That ChatGPT didn't make me anxious once I actually used it. It made me more curious. I started asking better questions about my own work. I realized I actually know more than I thought about certain things, and less than I should about others. And I realized that my value isn't in being fast at things—it's in being thoughtful about things that matter.
That's not a lesson I'd learn from reading about AI. I had to actually use it, mess up with it, see where it failed, and understand my own role better.
If I could tell anyone one thing about this: don't wait to be an expert to start using these tools. Just start. Use it badly. Make mistakes. See what actually helps you and what's just hype. Your instincts are probably better than the internet's panic.
Final Thoughts
Here's what I really think will happen in the next few years:
Some people will ignore ChatGPT entirely and become less competitive. Some people will panic, take expensive courses, and waste energy on a tool they don't actually need. Some people will use it thoughtlessly and make more mistakes faster. But most people—the people who are reading honest takes like this—will just quietly start using it and become a little bit better at their jobs.
You don't need to become an AI expert. You don't need to be afraid. You just need to be curious.
Try it for a week. Ask it to draft something you're writing. Ask it to explain something you're struggling with. Ask it to stress-test an idea. Notice what's useful and what's useless. Then decide what role it plays in your work.
That's it. That's the strategy.
I was wrong about ChatGPT being a threat. But I was also wrong about it being revolutionary. It's neither. It's just a tool. And tools don't determine careers—how you use them does.
You've got this. Now go try it.
Dattatray Dagale
Data Analyst • Blogger • Mumbai
I'm a data analyst from Kalyan, Maharashtra, working at Morningstar. I write about personal finance, career growth, and everyday life for Indian millennials — the stuff I wish someone had told me earlier.
Written by Dattatray Dagale • 30 August 2026
.gif)
.png)
.png)
0 Comments