A two-voice take on where to actually start. We've been having this conversation in a hundred shapes — over coffee, in war rooms, walking out of meetings. Here it is the way it sounds in our own words, side by side.
We keep landing in the same place, separately. So we wrote it down in our own voices, exactly the way each of us would say it.
I've been in this industry twenty years. In all of that time, there isn't a single tool that has stayed put. Rediffmail to Yahoo Mail to Hotmail to Gmail to WhatsApp. Man pages to Britannica to Wikipedia to Discord to an AI summary. One of the perils — and the beauty — of being in tech is that change is the baseline. I've come to read change as expansion. A better way of doing the same thing. You don't really get to opt out.
And yet the default mental model has barely moved. I work in a company. The company is supposed to tell me what to do, give me a budget, pay for my training, hand me the tools. That's a reasonable expectation. It's also slowly going out of date.
I didn't have a clean answer for this until I learned it from a senior teammate of mine. He said something very simple to his team:
It doesn't matter whether the company gives you the tool or not. A licence for any AI tool — ChatGPT, Claude, whichever — costs about ten dollars a month. Nothing stops you investing your own money.
Your own money. Rather than waiting for a training budget. That single sentence rearranged something in my head.
There are two parts to learning. Knowing is when you read about something. Integrating is when you bring it into your own life. Most of us are stuck at knowing. The fix is small and reversible — half an hour a day, or thirty minutes three times a week, on a tool or a skill or a corner of the stack you haven't touched. That's it. No grand programme. No new identity. Just thirty minutes that belong to your career, not the company's quarter.
One more thing worth saying out loud. Companies optimise for lowering risk. That's their job — for the institution. Your personal life doesn't have to inherit that posture. ngrok, for instance. Most large companies block it — for good reason, it's a known path for data exfiltration. When my Head of Engineering and I were debating a similar tool recently, I made the case the other way. I'd rather learn what it does, what the modern equivalents do, what the art of the possible looks like — so I can tell my team how to use it safely, when not to use it at all, and which guardrails matter. In the corporate posture, the answer to anything risky is don't use it. In your personal life, that posture costs you.
In the last fifteen or twenty days, on personal time, with personal money, on a personal laptop:
- A colour-picker tool for my partner Sripriya's customers, on sripriya.me — about eight focused hours.
- A site for the Conscious Leadership Circle I'm part of.
- This personal brand site, sreebalakrishnan.in. Every element my design and my thought — AI as the hands.
- An astrology companion — Node.js, role-based auth, the lot. Still cooking.
- chrisandsree.blog — a place Chris and I always said we'd write together.
ilearn.fun— a small sandbox where a handful of us play with tools that wouldn't fit inside the corporate fence.- A WhatsApp agent that listens to my messages and drafts replies for me. Not safe on a corporate device. Perfectly fine on my own.
What's underneath all of this isn't the projects. It's that I'm rewiring my brain. Some of it is curiosity. A lot of it is the quiet fear that, if I don't reinvent myself now, I'll be redundant. The cleanest part of it is the recognition that if I don't invest in me, I can't really expect anyone else to.
The way organisations have always absorbed new technology is, to my mind, fairly organic. Someone at the top decides — we're going to standardise on this IDE, this framework, this platform. The decision flows to Learning & Development. L&D and their technology counterparts work out a curriculum. Training programs are organised for the broader ecosystem — the QA, the SDET organisation, the platform teams — to start learning, testing, trying. At some point, months later, sometimes longer, it shows up inside your project lifecycle.
Now step into the new world of AI. Things started transpiring in the second half of last year. By December, they had catapulted to a different level. At first AI was a pair programmer. You reviewed every line. For a while it was a fairly poor pair programmer. Then sometime this year, it stopped being a pair at all — you give it the work, it raises a pull request, and you commit.
And the platform itself is changing under your hands every week. Open Cursor or Claude Code on a Tuesday and by Friday the version has shifted. New tools. New skills. Slash commands. Superpowers. Brainstorming modes you'd never heard of. It is information overload — and the only honest way to stay on top of it is to soak yourself in. To live the journey. I won't pretend it isn't overwhelming. It is. That is the reality.
We are, as an organisation, a chariot pulled by many horses. If one horse pulls in a different direction, the whole chariot slows. So one thing I keep emphasising: please don't wait for someone to come and hand you the framework, the guideline, the tool. That hand-off is a function of how many people are available to nurture you, how quickly the org can stand up licences, how procurement calendars align with the technology one. It will come. But it will come late.
Treat it as an essential investment in your own development and your own capabilities. I openly encourage everyone to think of it as going back to school. When you're a twenty-year-old in school, nobody is talking about work-life patterns, or how many hours you put in, or whether you should be working on a Saturday. I'm not advocating toxicity. I'm not asking anyone to push themselves into something unsustainable. I'm talking about a posture of innovation and curiosity.
One important caveat, and I want to be clear about this. It cannot happen at the cost of your work. You are learning and relearning new things, and trying to do that during the working day will compromise both. Spend your weekends — even four hours on a Saturday, four hours on a Sunday. Or spend a focused hour every day. On your own laptop. With your own licence. Your own model. If you can't afford the paid one, go open source. Experiment. Break things. Try.
Expecting the company to put it all on a silver platter — the hardware, the software, the curriculum — is wishful thinking. Don't wait for that time. This is the right time. Start investing your time. Time, unlike many other resources, is not infinite. And if making your time worthwhile requires you to spend on certain hardware or tools, spend it. For your own benefit.
There will come a moment when the company catches up — when the licences arrive on the right timeline, when the assets are placed in your hands. When that happens, you'll be in a double-garden. You're already in good ground. Now you simply start compounding on top of the new facilities. That's the path.
We keep landing on the same line, separately and together:
AI is not going to replace you. Someone who knows how to use AI will.
This corner is where we'll keep thinking out loud about that — about delivery, about teams, about what the practice of building actually looks like in 2026. In two voices, on the same page.
— Venky & Sree