And Vijay Says…

As I am building out my agentic AI team, about half my time is being spent on hiring great technical talent. I thought it’s a good idea to explain what I look for at a high level for the four roles that I am hiring the most

Hopefully this gives you an idea of what I look for at a high level in each role. It’s not an exhaustive list. I am not listing basic competence needed in each role at all – someone else will usually evaluate all that before you and I will have a chat.

I hope this helps you qualify if you are a good fit for these roles. If you do think you are a fit – we would love to talk to you. It’s a great team and we are fun to hang with and we do a lot of cool stuff that the biggest companies in the world benefit from.

Great engineers

They are great craftsmen – not just creating code that works, but code that lasts. They understand that their work doesn’t exist in isolation – others need to work with it. While they are curious about technology, they are pragmatists. For example – they are not religious about always designing for horizontal scaling and will happily start with vertical scaling that is simpler and faster if that’s all the problem needs for a good solution.

They write good documentation that lets others understand their rationale. They don’t wait for QA to find issues – they take pride in testing and spend the time thinking about corner cases.

Great engineering managers

There was a point in time when I was a half decent engineer (maybe with the exception of writing great documentation) . I struggled as a leader of an engineering team though. Later in my career I observed that this was quite common – and it’s no different in sales. Great sales people don’t automatically became great sales leaders, and great sales leaders don’t automatically turn into great P&L leaders.

It’s a hard transition when you shift focus from managing code to managing people who often are way smarter than you.

If any role in a technical team needs to be good at politics – engineering managers are the ones that need it. They need to be able to find resources for their team to succeed and shield their team from EM organizational craziness. There were days in my life where I truly just wanted to go back to being a technical hero and not deal with product managers and sales people and CFOs.

EM needs more EQ than IQ. Great engineers are a diverse lot – some need to be left alone, some need active coaching, and some will burn out if you don’t keep checking in to course correct. It’s a tiring job and you need to thrive in that environment

They need to be communication wizards – negotiating with product managers to translate the roadmap into things that can be coded, while also explaining to finance why you need budget to tackle technical debt.

I have always felt that it’s easier for an EM to learn basic business than to teach tech to other stakeholders. I have often felt my MBA was wasted effort but grudgingly I will accept that it helped me explain tech better to others in a language they can follow.

My boss BK keeps telling us “candor over comfort” is how we should operate. This is critical for EM . There are very few days without hard conversations in this role

Great architects

A few technologists need to step back and use wide angles lens to look at problem statements and solutions – that’s the architecture profession.

There is no such thing as perfection in architecture- it’s a series of tradeoffs. You have to balance what is needed today vs long term sustainability. Architects are not philosophers – they have to weight pros and cons and come to a decision. Poor architects are quite easy to spot – they throw around a lot of jargon and never take a hard decision. Great architects are harder to find easily – their work looks boring because everything just works.

No one can predict the exact future – architects are not prophets. But what they need to do is to make sure that software is written in such a way that the next requirement won’t force a complete rewrite.

Unlike EM – architects don’t usually have big teams and often are individual contributors. They have very limited direct authority. So they need to influence decisions via their clarity, with prototypes and with solid technical rationale. Architects who preach from an ivory tower lose the respect of engineers really quickly.

Architects need a detailed understanding of the business problem they are solving – the job is to identify capabilities and design solutions for them that accelerate the delivery of those to the business.

Great product managers

Some of my PM friends don’t like me saying this – but I like PMs to be the CE”No” of the team. They have to say NO to a lot of things and ruthlessly prioritize.

While engineers and architects are tasked to build the solution the right way, product managers are the ones making sure that it’s the right solution that is being built. They are the connective tissue that makes it all work. Engineers take care of the HOW – PM takes care of the WHAT and WHY.

Saying NO to powerful people js not easy. You need clear frameworks to make decision – and those frameworks need to be data driven. That in turn means you need to understand what adds value to your clients and what moves the needle for your business.

The viability of a product is not just determined by the coolness of technology used to build it. Yes agentic AI is super cool and you need to know how it works – but even more important is understanding how the intersection of UX, regulations, GTM, pricing etc work

Much like I mentioned about architects – this is a role that is low on authority because they don’t manage big teams. They influence with clear thinking and market insights.

As we grow into more senior roles, constant fatigue becomes a real risk for most executives. I have felt this myself and I know several friends who are absolutely miserable. Very few talk about it openly. In my friend’s circle, this has been a recurring topic for a while. We joke about it as the natural outcome of getting old – but we are all well aware that age is not the root cause, or at least not the biggest reason.

I am in a plane now on a work trip to India and thought I will post a rant on this topic.

As I look back at my own career – I can see a certain paradox. I have been exhausted from having too much to do. I have also been equally exhausted from having too little to do. Either way the outcome is terrible. Exhaustion should not be worn as a badge of honor if it’s perpetual.

In my case – being bored out was an even bigger challenge. My brain just switches to a low power mode and I kept doing stupid repetitive things mindlessly. I grew bitter and there were physical symptoms like throwing up and so on. Thankfully I could change roles relatively quickly when I ran into those situations. I no longer accept roles unless I am very sure that the mission excites me. It took some hard knocks before this wisdom kicked in and I wish I was smarter about it earlier.

These days I ask myself a simple question “What will break if I am not around for a couple of weeks?”. If not a lot will break is the honest answer – it is usually a good indicator that being bored out is just around the corner and I am largely doing repetitive work that can be delegated or automated. Then it’s time to find a new mission. You and your boss might not agree about this though. Be fair and give your boss the time and context to process this. Also the mission might not be a new job or role – it could be a side project that excites you and is white space for the company.

Since I have also worked with and managed a lot of execs over the years whose careers I have observed closely, I am reasonably sure this has been the case for at least a good proportion of them.

Most senior leaders walk unsuspecting into an agency trap. The transition from an individual contributor to a manager is hard. It’s equally hard if not harder to become the manager of managers and every step up from there. It takes some time to realize that you no longer can directly influence the outcome like you used to in your previous role. I still struggle with this every time I take on a new role, but usually get over it faster thanks to the life experience gathered along the way.

As I look back at the relatively low exhaustion roles I have had earlier in my career – I think one reason is that I had a lot of peers I could freely talk to. I didn’t have to think about “is it ok to be vulnerable?”. Every step up after that, my peer group shrank and it took me a while to realize that it’s ok to be vulnerable to your team as well. I can’t say I have quite mastered it but I am a lot better at it today than say 15 years ago.

Senior leaders are managing the collective anxiety of a lot of people in their organizations – and often outside their organizations. You need some venting mechanisms built quickly to deal with it. The comfortable thing to do is to lean on others in the same situation – your peers and your direct reports and your boss. There is value in that – but there is a challenge that we might not immediately realize. It’s super easy to fall into an echo chamber where everyone feeds into the insecurities of everyone else.

Echo chambers are a reality. I don’t think there is a way to avoid them for good. The closest to a good strategy that has worked for me is to spend time with people with very different interests. I find time to hang with people who train dogs (they have their own echo chambers but it’s so different from my line of work and it’s almost therapeutic for me to deal with that for some time). I am an avid reader – and I consciously spend time choosing books outside the world of technology for about half my reading. It works – it not only reduces my cognitive load, it even sparks new ideas for my line of work

Somewhere along the way I lost interest in chasing vertical growth in my career. I started enjoying helping my team grow more than my own progression in the hierarchy. It didn’t stop me from getting promoted multiple times after that – that’s the irony of this story. It is quite a liberating feeling when you don’t feel like you need to run at an unsustainable pace. It’s not like I don’t care for money and some luxury anymore – I am not a saint by any stretch. It’s just that I value other things also in equal measure and can make better trade offs.

One of the things I try to do is to take a pause and do an ROI evaluation on the important things I have been doing in the recent past. I make a list of things and people that give me energy and another list of things and people that drain my energy. Then I find ways to spend more time on the former and less on the latter. I don’t always succeed but even if I can shift the balance by ten or twenty percent, the results are amazing.

This is the report that wiped out $200B of market cap the following day. https://www.citriniresearch.com/p/2028gic

First – I don’t blame the authors one bit. They painted a scenario – they didn’t predict that’s what will happen. CEOs of the frontier labs and hyperscalers have all painted their scenarios and we were all cool with that. So why not this one and more like it?

Second – I absolutely blame economists and professional investors for being stuck in the past. They have been lazy doing just the first order impact analysis of AI ( picks and shovels will be of great demand in a gold rush so buy Nvidia stock ). That doesn’t even justify their plane tickets to go to Davos and pontificate on AI annually at WEF. Their lack of rigor in their field is what I attribute the current volatility to the most.

I am just an engineer who plays with AI. I have some experience running parts of the P&L for larger companies. I am not an economist by any stretch of imagination. I am going to jot down my thoughts on the scenario painted by the report mostly so that I can come back in 2028 and read about it. So please take it for what it is worth.

As always, these are just my personal opinions.

I will start with the conclusion and explain why I think so across a few bullets. I think the chance of the economy crashing in 2028 because of AI is completely improbable.

Here are 6 reasons why

1. History has always rhymed – every single technology shift in the market has led to more employment and the kind of roles it creates were not known when the tech first came around. That has been the case for Steam engines to ATM to internet.
2. There will be a market driven spending brake. If there are no people to buy products, there won’t be runaway AI investment either. This idea of firing people and investing in AI is not an endless loop like the report says – there is no economic logic that supports it.
3. If and when AI makes everything from food and housing to sneakers cheaper – workers won’t need to make as much money to keep their current standards of living. A four day workweek is a likely outcome in 2028 than economic collapse
4. Just look at public cloud adoption in large companies to get a hint on how regulations and bureaucracy slows down tech by decades. On top of that there are labor unions, courts etc as well that will cause further friction that slows down adoption. The talk on AI is miles ahead today compared to the walk of AI. The layoffs we see are largely from past over hiring – not because AI has taken over jobs yet. It’s just a convenient excuse for companies to shed cost now.
5. White collar workers – not all but many – have access to capital market directly or through their 401(k), IRA etc. The wages they might lose in the doomsday scenario will easily be offset by the capital appreciation they get from Nvidia etc booming if the scenario indeed plays out with the Ghost GDP. Cash is fungible – doesn’t matter a lot which way we earn it . It matters for tax etc but you get the point
6. How likely is financial contagion? If it’s Amex and visa getting disrupted by AI agents who prefer solana blockchain – there are plenty of regulations around KYC/AML that make it happen. Mortgage crisis is a real potential in my mind – it can only be partially offset by the access to capital market that some of the mortgage holders have. But because of the spending brake I discussed above – I don’t see a full blown disaster

So back to who is to blame for this volatility.

Economists need to change their 20th century ways of thinking and start modeling the future. I would like to think the good ones have already started and perhaps just were late in starting. It makes me wonder if economics education model itself needs urgent change. The world we live in is moving so fast and picking up speed that we can no longer just be happy with first order impacts modeled.

Same goes for professional investors. You are playing with the money of people who trust your knowledge and competency. If one doomsday scenario paper is enough to spook you to act this way – you need to reevaluate your investment thesis from first principles

AI may or may not change the economy drastically – but I sure hope it does change the practice of economics and professional investing for the better.