How to 10x Your Value in the A.I. Era | ft. Kunal Shah
Groww India Investor Festival 2026. CRED 创始人、WhatsApp 印度负责人 Kunal Shah 与 ETtech 编辑 Samidha Sharma 的深度对话。
中文摘要
Kunal Shah 在 Groww 印度投资者节上对 AI 时代的财富创造、高能动性（high agency）和职业生涯给出了一系列尖锐判断：
- "世界最大雇主是无效率"——AI 取代的是 inefficiency，不是 abstract "jobs"。收入差距将在使用 AI 和不使用 AI 的人群之间急剧扩大，而非简单的"就业 vs 失业"
- 入门级陷阱：初级岗位（可外包、任务导向的工作）最先被 AI 替代。"外包"这个词正在被"outsource to agents"取代。没有入门级岗位就没有中层管理者培养路径
- CRED 内部：AI 写代码比例从 5%→90% 仅用一年——这是 Shah 自己的公司的数据
- 10x 工程师现象：每个技术公司 top 10% 的员工正在变成"完全不同的物种"，他们开始要求去掉会议、不管理下属、连续工作 5 天然后休 5 天
- "大公司就是印度交通"：如果把 10% 的车升级到 1000km/h 但放在同样的拥堵道路上——你不会看到系统提速。大公司处于巨大劣势
- Apex Predator（顶级捕食者）隐喻：变代谢率、高捕食成功率、极端适应力——鲨鱼/鳄鱼/螃蟹这三类存活了 1 亿年的物种的共同特征
- 印度特定问题：35GB/月人均数据消耗（全球最高之一）、"效率/生产力"在大多数印度语言中无对应词汇、女性劳动参与率低、过度依赖政府/公司解决问题的心态
- 个人行动纲领：把自己当股票看（提高增长率）、把自己当 app 看（持续迭代）、技术跟年轻人学、价值观跟老年人学
Introduction [00:00:00]
Samidha Sharma: Thank you everybody for coming today. The markets didn't give me the confidence that I'm going to see so many people here but we have with us Kunal Shah and we're discussing wealth creation in the age of AI — who wins and who doesn't. I've always felt that Kunal is somebody who has very sharp understanding and takes on mostly everything that concerns any of us.
Kunal, I want to just start — I'm assuming most of the people here are already in the workforce or are prepping to be mid-career professionals. This is a super relevant topic. It ties back to the entire discussion on jobs and AI eventually. Your jobs take you to creating wealth. Where are you at when it comes to jobs and AI? There are two schools of thought. One which says this entire AI apocalypse is all fantasy, and then we see what we are seeing with all of these big companies cutting jobs. Address the elephant in the room.
AI and Job Concerns [00:01:26]
Kunal Shah: First of all, how many of you use any of the apps like ChatGPT or Gemini every day? [many hands] Awesome. And how many of you have built one agent of your own so far? [fewer hands] How many of you are concerned that you're going to lose your job to AI? [not many] How many of you think that your colleagues are going to lose their job to AI? [more hands] Awesome. I think this is a better understanding of what's really happening.
I'm old enough to tell you that I've lived a time where internet was not there in most workplaces or computers were not there. When I went to my first job there were no computers. At that time, internet made information accessible in the air. Now you could just search anything and get anything. AI has made intelligence accessible in the air. This is a very hard to understand concept because intelligence was a unique trait that we attributed to humans as a species — and now you can just plug it up in the air and just suck it in, and you can actually accelerate materially.
AI Displacing Global Inefficiency [00:03:35]
The second thing is that we forget that the largest employer of the world is actually inefficiency. At the pace inefficiency is going away, the current jobs are getting displaced or changed materially.
A lot of people worry about job layoffs. I think the thing we have to worry about is per capita income not growing. There are enough people in India who are still doing liftman job and ATM guard job — they all have jobs. There are rickshaw drivers who are currently operating without maps and still on the street — they make less money than the people who have apps on the phone, but they exist.
The thing is that the income delta between people who use technology and use AI versus people who don't use AI is going to be wide. The world has interesting ways to kind of keep people employed. But the real question we have to worry about is the per capita income growth.
A lot of us are waiting for our companies to teach us AI or some workshops to teach us AI or show up at this event and hopefully learn something. The truth is everything is up there. If you are really keen to learn, you will figure out everything on your own. There is a huge shortage of self-motivation, self-discipline to get things done because now every single thing is transparently available. You can build anything, you can create anything, you can research anything. The question is how many people are actually using these tools for better stuff. A bulk of people are still using AI for frivolous use cases and not for what it is truly capable of.
Frivolous AI Use Cases [00:06:10]
Samidha: What do you mean frivolous use cases?
Kunal: Well, we are just still editing pictures and videos, talking to ChatGPT like a friend, and getting rid of our real friends. [audience laughs] How many of you talk to ChatGPT as a friend? How many of you have friends who are talking to ChatGPT as a friend? How many have done astrology on ChatGPT? How many of you have done therapy on ChatGPT?
If we're talking about wealth — the most valuable wealth you have is time. And what you do with time becomes even more important than it was before. India is now at 35GB per capita usage per month — that's much more than any other country in the world. And 35GB usage requires a lot of time to be wasted. Short videos and time pass and all of that stuff.
Elon Musk, Sam, Kunal — everybody has only 24 hours. What we do with that time is becoming even more important for wealth to happen. It has become extremely easy to also waste time in the world we live in. We have the same device that can make us super intelligent or super dumb very fast — within weeks you can actually see yourself go in opposite directions.
India's Productivity Culture [00:08:04]
My worry right now is productivity as a concept. The words efficiency and productivity do not exist in most Indian languages. That vocabulary is not in our core DNA. We don't value time as much. There is no word for efficiency and productivity in any Indian language. Most of you can ChatGPT and find the answer but you'll find that they are actually surrogate words, not actual words — because productivity and efficiency words came from the Industrial Revolution where time became valuable. We are still to realize that. Most jobs in the Western world are hourly jobs. We have never been paid on an hourly basis, so our brain doesn't work in hourly mode.
AI Wealth Creation Gap [00:08:46]
Samidha: What does all of this mean when you look at the world right now? My observation is that it's literally 1% of the population which is really aware about how stark the chasm is between what we know and what 99% people across the world are aware of in terms of what AI is going to do to their jobs and wealth creation. This one constituency in Silicon Valley — just like how the gold rush happened in California — they are so aware, they are the beneficiaries of everything that's happening around AI, and then there is the rest of the world. How do we keep up and even get to partake in that wealth creation?
Kunal: I would say China is not too far away. The gap is getting wider and wider in terms of just the sheer awareness. It's become hard to even keep up on a weekly basis what's happening to the world. When I meet people socially, what they talk about AI seems so ancient because it was something from last year — and last year feels like ancient time now.
India has a very unique tendency when it comes to these situations. Remember COVID? We had only two modes — apathy and panic. There was nothing in the middle. We didn't give a damn and then we panicked like crazy. I feel we are going to go through the same phase with AI. Right now we are in a phase of apathy because it doesn't seem to be impacting us. But as soon as it gets close and there are second and third order effects...
For example, let's take a bank. Banks today or financial services form 30 to 40% of India's market cap. In a bank, a lot of stability comes from lending, and from lending it comes from IT and BPO jobs — which would form 30 to 40% of the bank's book. Now even if 10 to 15 to 20% of jobs get impacted, the most safe part of the book starts getting impacted — which means the bank's book starts getting negatively impacted — which means 30-40% of India's market cap. I don't think anyone's talking about that.
The Entry-Level Trap [00:11:58]
Samidha: A lot of the jobs that were outsourced to India are actually significantly more likely to get impacted versus the jobs that were not outsourced to India — because their complexities were different and therefore you chose to delegate work. What you would assign a junior person to do now can be done through ChatGPT or any of the tools. The gap is getting wider. To be eligible for a job, college graduates are going to have an even bigger struggle than they did before. You'll wonder how many years you have to keep studying to keep being eligible for a job. That's where judgment comes in.
Between the ages of 20 and 25 is actually when you hone skills to become a middle manager, then you rise up the ranks. If you don't get those people into the workforce, how do they even get to the second and third stage?
Kunal: We're talking about this today in May 2026 when AI is evolving exponentially every month. By the time we talk about this in a year's time from now, the issue may get even more stark. The issue is: who's supposed to fix it? We have gotten addicted to somebody else solving this for us — that government will come and solve this, or my company will come and solve this, or my parents will come and solve this. A large majority of Indians are still doing arranged marriages.
We are in a time where everybody's on their own and they need to operate with very high agency — solve for themselves, become very very good at things, because now the time you need to become good at something has completely shrunk through all of these tools. It's just: are you doing something about your time or not? It's also the time you can waste your time the most.
The reason I worry about India specifically is that when you have such large data usage, when you have a large young population, when you have jobs that can potentially get impacted through any of these tools at entry level — we are looking at the impact to get more severe unless we create a mission-oriented mindset saying we need to ride this wave.
Future of Outsourcing [00:14:42]
Kunal: I used to run a BPO company before — a large outsourcing setup out of India for US companies. I remember this line that the American community used to say: "Do what you do the best and outsource the rest." The word 'outsource' will get replaced to agents. Outsource will get replaced to AI. Outsource will get replaced to robots. People will keep doing the best stuff themselves.
The question is: have we built resilient industries and capabilities that can survive? We have multiple issues to start with. Female per capita growth is not happening. Our per capita income is lagging because female participation in labor is not to the level the world expects us to have. What percent of loan takers are women? My guess would be sub 10%. What percent for credit cards? Sub 20%. How do you expect the country to get richer with one gender working? It's never happened — unless you find oil, which will magically make all of us rich. But it's unlikely to happen.
Building Real Substance [00:16:36]
Samidha: What are three actionable things a 24-year-old should be doing right now?
Kunal: First of all, they have to accept that they cannot say "I'm not very good with tech." There is not a single human being that can afford to say "I'm not very good with tech" — because every job is.
[Samidha pushes back on inequality of access]
Kunal: I disagree with that completely. We have the highest smartphone penetration. We have the cheapest data. If you can use something to waste your time on short videos, you can also spend some time learning some tutorials on the internet. We've gotten used to the excuses all the time.
We have to truly understand and become significantly more ambitious than we are. The other issue is that we are looking for shortcuts to get wealthy versus really building substance in ourselves to become wealthy. If everybody thinks there is a lottery out there — and I'm speaking at a Groww event, it sounds wrong when I say this — the mathematical odds of getting wealthy through investing apps are absolutely against you. You have to build substance in yourself.
All of us have built a little bit more gyan before we build substance. We don't understand things deeply but we have opinions about everything. It's become extremely easy to have opinions about things right now. It's very hard to build substance because substance requires you to have sacrifice — and sacrifice has become a harder problem for us.
It's controversial — I know Narayana Murthy got a lot of flak for saying we're not working hard enough. But how can a poor nation afford to work 5 days a week? We compare ourselves to the Western world and say we need to have 5 days a week, but we also have 25 holidays which the Western world doesn't have. We live in this convenient reality but the world is moving much faster. People's incomes are growing faster. Their per capita income growth is happening.
Embracing Risk and Struggle [00:20:02]
Kunal: We are going to definitely have to let people go through pain, and that's when any country evolves. When Japan took off, it was after a major setback. Constantly protecting people from setbacks and reality has stopped the progress for the country. Why do successful people struggle to have kids who can be successful? Because they take away the exact same thing that made them successful — which is struggle. And that's exactly what we're doing to more and more people.
60-70% of population is already in some protection. But for the rest of the 20-30% — especially the educated cohort, especially people who have access to all these tools and the privilege — they need to take a lot more risk.
I also believe AI is a time where we will truly become 10x. Every single person over here has the opportunity to increase their income by 10 times. It was never possible before. This is really true now. It's just: how do you apply it and how do you meet such people?
You are the average of the five people you spend time with. India has another unique problem — we find it very hard to let go of people who are not growing. We are a very social society where it is considered a bad thing to move on for better choices in terms of who we spend time with. But the truth is: if you feel bad about leaving the five people you spend time with to grow, that's going to be limiting your own growth — and nobody can be blamed for that except you.
Corporate AI Preparedness [00:24:05]
Samidha: How prepared are companies? We're seeing a lot of news coming out of the US — 10%, 50% — actually the biggest tech companies which are deploying crazy amounts of capital in AI are also shedding workforce. Very very profitable companies like Alphabet, Meta. FII outflows have been record high, and one theory is that there is no AI play for India. Where there is no big lab or foundation lab like OpenAI or Anthropic, how does the workforce suddenly independently become very AI native?
Kunal: It's a little bit like a game theory problem. Let's take banks. If one bank decides to really accelerate on AI, the real metric that will move is revenue per employee — it will go through the roof. Today, revenue per employee is very similar across all banks. As soon as AI becomes important, the leverage will shoot up. If Bank A's revenue per employee becomes 2x, suddenly they will dislodge all the market shares on everything else. That's exactly what happened when computerization came or internet banking came — one bank took the leap and suddenly everybody else was forced.
What percentage of India's public market profit pools are protected by some regulation, license, or public government companies? A very large chunk. So the impact may not be felt as quickly in the sectors that are protected. But we will not be globally competitive anymore unless we are only dependent on the local market. When it comes to global maxima, AI will matter. When it comes to local maxima, it will not matter as long as there is protection through regulation.
Rise of 10x Engineers [00:27:39]
Kunal: 90% of code at CRED is written by AI right now. This number would have been maybe 5% a year ago. 5% to 90% in a year's time is an absolutely crazy jump.
10% of every tech startup's employees are becoming a completely different species from the 90%. Their productivity is now at exponential scale, and they are actually finding every single person around them and every other process to be slow. If you want to build an org around those people — number one, or you'll be disrupted by people who are built by only those people.
We used to make jokes about "what is a 10x engineer." Now you see it all the time. Very fascinating: when I speak to these 10x people and ask them what can I do to make your life better, they're saying: "Can you remove my meetings? Can I stop dealing with other people in my team? Can I stop having people report to me? I'll do this work myself. Can I not talk to my manager? Can I just work for 5 days non-stop and not work for 5 days? I'll get 3x more work done." These are conversations I've not heard before. These are all real conversations right now.
Samidha: How much of this is translating into output being better? There are companies still not sure if the output is as good as the kind of monies they're putting into Anthropic or OpenAI.
Kunal: It's a little bit like the India traffic situation. If you only make 10% of the cars capable of running at 1,000 km/h but they are kept in the same traffic where every other car is operating at 10 km/h, then you will say traffic has not moved fast. Companies that are only hiring 1,000 km/h cars are moving much faster. If you keep very high displacement people with low displacement people and low displacement processes, you will not see the impact. Large companies are at a massive disadvantage right now because they're exactly like the India traffic situation.
Final Career Advice [00:30:38]
Kunal: A few things I advise people in their mid-20s:
-
This is the best time to become extraordinary at learning anything in a very short period of time. If anybody tells you "I'll take three years to learn something" — they are anyways not going to make it. You have to assume that any topic, any complex topic, you should be able to learn in days and weeks. I'm not saying you'll become a doctor in two weeks. I'm very specifically talking about very specific skills.
-
Who are you surrounded by? You, your average of five, is a real thing. Think of people around you as stocks — and they have a growth rate. If their XIRR is not high, your XIRR will not be high as well.
-
Think of yourself as a stock. What is your growth rate? Nobody's coming to solve your growth rate — you have to solve it for yourself. As an investor, do you go and help the companies you've invested in? No. Same way, nobody's coming to help you. So increase your own growth rate.
-
Think of yourself as an app. If you have bugs, fix it. If you have features, enhance it. But keep releasing more often. Keep evolving. If you're going to stay the same, the world is not going to be waiting for you.
-
You are better off learning from younger people when it comes to technology. You are better off learning from older people when it comes to values. Don't go to the wrong places for wrong things.
-
Apex Predator mindset — I researched which species survived 100 million+ years without changing too much: crocodiles, sharks, and some crabs. Three common traits:
- Ability to change metabolism at will — can be at 10% energy or 100% energy as needed
- Very high strike rate — when they attempt to get food, very high conversion. Deadliest bites. Mostly chilling otherwise
- Extreme adaptation — can survive ice age, climatic catastrophes
If you want to be an apex predator in the AI era, follow these exact same things. If you struggle with any of these three, you will struggle. Metabolism in this world means: "Oh, my capacity is only this much" — you have to be able to accept that you'll have one-tenth of the capacity and still survive and actually crush it. High conversion rate means when you choose to attempt something, absolutely grab the opportunity. Don't waste time at events. There's no time left, guys.
关键引用
"The largest employer of the world is actually inefficiency."
"AI has made intelligence accessible in the air. Intelligence was a unique trait that we attributed to humans as a species — and now you can just plug it up in the air and just suck it in."
"The income delta between people who use technology and use AI versus people who don't use AI is going to be wide."
"90% of code at CRED is written by AI right now. This number would have been maybe 5% a year ago."
"10% of every tech startup's employees are becoming a completely different species from the 90%."
"The word 'outsource' will get replaced to agents. Outsource will get replaced to AI. Outsource will get replaced to robots."
"Every single person over here has the opportunity to increase their income by 10 times. It was never possible before."
"Large companies are at a massive disadvantage right now because they're exactly like the India traffic situation — 10% of cars running at 1,000 km/h, 90% at 10 km/h."