Product managers is not just about making product features. It’s a job that requires lots of human interaction, and this part is notoriously complicated for engineers. Article states that a person coding the feature can build a product intuition, and that’s true, product is not sorcery, but from my experience most engineers are just not interested in that. That may change, of course, if engineers will be pressured by economic situation, but until that they’re happy when someone else is interacting with users and stakeholders.
at my company they are trying to get us engineers to absorb that role so now we spend a lot of our time doing meetings and human interactions like this with other parts of the business and making decisions
idk how common this is in the industry though, but since the introduction of AI my role has expanded so much beyond what is just writing code, i'm now a devops/product/dev person and expected to do all of that at once for the same pay as before
i actually enjoy it myself cause you feel less like a ticket pusher
It is common, and you can resist (risking your salary), going deeper (hardening engineering skills) or wider (doing product, analytical, etc. work). I believe that an engineer with product mindset is extremely valuable for a company, so I would encourage taking that route. However I’m also seeing engineers taking first or second path.
As long as humans are making decisions, human interaction remains important. From this perspective, PMs are still in a better position than engineers
I'm a software engineer and never had a job with a dedicated product manager for the first decade of my career. I hate to use the E word here ("empathy") because people say it performatively too often for my taste but to me it's the key to success in this capacity. People will tell you a lot if you listen and choose to engage with them on a passionate level. Sometimes their feedback will be useful and correct and sometimes it'll be a rant fueled by something you have nothing to do with. Either way you'll gain insight into how the thing that you're building is used. That's very valuable information. I've worked places where these conversations didn't happen and it wasn't to my liking.
You're right that this is going to suck for the engineers who fit the stereotype, didn't like interacting with users, and would express that they just want to stay in the code.
We don't need code anymore, we just need intermediate deterministic artifacts that the AI can produce and the user can validate in order to circumvent the - by definition - non-deterministic nature of LLMs.
What should we call these intermediate deterministic artifacts? Code!
I think there's a big distinction between needing code - the machine abstraction for electrical logic gates - and code that will ever need to be read by a human.
Yeah we'll still need to know a sequence to flip bits in the right order, but it isn't code you'll recognize.
You fail to appreciate that creativity is more than just code. I would argue code is the easy part. The difficult part which you obviously dont appreciate or value is being creative and "different". Thinking what or how something can change or be better.
You're missing the point about how the industry is going to go through a great narrowing. In the past, you'd have one PM and 8-12 engineers. Now you only need 2-3 people with both product and engineering ("architecture", to quote the author) skillsets. You are absolutely correct that most engineers are not interested in the Product side - those are the engineers who are going to get cut in this great narrowing. Quoting the author:
> I grew up with the distinction of "tech guy" needs a "product guy". I am not dismissing it, I am just saying that the world is entering into a phase where both roles are eventually melting into a single function.
The whole reason why Engineering exists in organizations is to take Product's ideas and turn them into reality. When Product can realize its own ideas (through AI) then Engineering is no longer necessary as a political sub-organization. Making it happen requires engineers to learn Product skills and adopt a "Product" job title. We will likely call people with both skillsets Product Engineers. People who can't adapt to learning both skillsets are at risk of losing their jobs.
But “product” people can’t really build real production software even with claude. Prototypes, sure.
Engineers can (and many do already) deal with product topics. It’s easier for engineers to become more produc oriented than for PM to be able to handle tech topics with AI
Gen AI has changed every single part of classic software development, and the impact hasn't even remotely been felt. And zero jobs are "safe" in that traditional way that we saw them.
It's actually staggering to think of how the teams I've led developed software five years ago, ten years ago, fifteen years ago and twenty years ago. Easily 95% of that effort has been obliterated (I'm estimating on the low side, more like 99.5%). Classic wisdom like "the specification needs to be 100% up front because once we implement it stratifies" are relics of the past.
Everything has changed. I can't fully express how software development will work, but it certainly won't involve someone whose job is "human interaction".
I don't believe other jobs are going to be as impacted as software engineering. Software is unique in that regard that the results can be verified quickly and easily fixed. On the other hand nobody is going to try building a building only to check if it's structurally sound because that would be absurd.
A sufficiently sandboxed, low-stakes thing that self-verifies by its visual output can be easily created by AI.
For security or applying business logic (and how it interfaces with programming or database logic), you get less and less assurance by just looking at it.
> Software is unique in that regard that the results can be verified quickly and easily fixed
Neither of these statements are true if your team is building anything remotely complex with real customers. You can’t quickly verify if your decisions will survive the test of time and past mistakes can’t always be easily fixed. LLMs trivialize the easy parts of software development and enable faster enshittification of the projects they’re used on, but that’s about it.
Paradoxically, gen AI being good at software engineering massively accelerated its impact on many other fields as a second order effect. Once generative AI could build code to analyze data, etc, its reach massively expanded.
> On the other hand nobody is going to try building a building only to check if it's structurally sound because that would be absurd.
Structural engineering isn't some mystical science. It's a set of mathematical truths, calculations, SOPs, and so on. If a human can do it, of course a machine can do it, and I'm obviously not saying "run it through the stochastic parrot and out comes a corporate headquarter building", but it's a part of the process where we're rapidly going to get there. And of course, the disruption in that field is already underway. For instance Stru AI. Yes, an engineer is still involved, but the premise is that your single engineer now replaces 10 engineers, and so on.
Gen AI isn't going to be plunging toilets or changing bedpans...yet...but it disrupts almost every role in the "white collar" world. That doesn't mean it eliminates every job, but it completely upends the norms.
PMs ultimately take bets. Some to many will fail and if they are great a few will pay off in an excessive way. I have yet to see anything in the AI field that make better bets than my best PMs (I manage 50 of them) but as soon as I do the whole calculus changes. I joke with them about it. They also joke back it could happen to me (which it absolutely could). We live in interesting times.
You seriously manage 50 human PMs? That sounds really interesting, I can't even imagine what kind of role that would be, other than someone running multiple companies simultaneously. Could you share what you do?
No, a very large software company. Have a team structure so not dealing with day to day but ultimately responsible. Have spent last year blending the team with engineers who have been interested in getting into product. Ultimately see that will be direction of travel across the industry.
Of course PMs aren't safe, but the article makes a great point why they're currently more relevant than ever!
AI can now push new features at speed, but somebody still needs to steer those features into a desirable product. Customers will ask AI for a faster horse.
This is why AI companies aren't spinning out new companies left and right. They can obviously push huge volumes of code, but can they effectively manage the resulting products?
Huge volumes of buggy code. The ChatGPT web interface should be an absolute embarrassment. Tap the input box, scrolls to top. Submit input, scrolls to top. 1+ second delay from keypress to input. Scrolling locks up. It’s been an absolute buggy mess for ages, the specific bugs change day to day. I suppose this is what happens when people think the LLMs are good enough to let loose without supervision.
IMO if developers are “cooked” like you stated (would be nice to have more data tho) then certainly plenty of other white collar jobs are in even worse position. They are in a bit better position than many other corporate jobs when it comes to transfer of skills and learning new profession.
> Code spitters are cooked, let's be honest here and accept the fate so everyone can move on.
I'm guessing "code spitters" here means people who consider their job to be manually writing code based on requirements produced by others.
Personally, I think the role of "developer" has already expanded way beyond that. If you're a technology expert who understands what is and is not possible to build, how it can be built, how to prompt that thing into existence, verify that the implementation is good, then get that implementation safely into production, you're not "cooked" at all.
For once I actually agree with Simon, and fortunately the percentage of engineers interested in software architecture and business concerns is so low already that I doubt they’ll have trouble in the job market. Pure “coders” were not safe from layoffs and PIPs even before LLMs, in my experience, since their jobs could so easily be outsourced and offshored.
Is it really that grim? AI helps a lot and is great at certain well defined tasks, but the “AI team” at my small org launched a new product very fast and very buggy. Customers were pissed. They had to go back to reading the code and applying some thinking and care to get a working product out. Still faster don’t get me wrong but there’s still a lot of tradeoffs to be made that AI can’t really do alone. Like there’s no objective correct answer.
I think it is: From my personal experience, it's the orgs that fail much more so than the technology. If you have a few capable engineers with AI that are able to think strategically and prod the agents from the right angle, you can move ridiculously fast.
Code quality improves month by month, and by now I rarely have complaints in PR reviews, even if I read through them manually, line by line. For the most part, it's how I would write it myself, only that I would have been slower, and not as motivated to test it so thoroughly.
The key is IMHO giving the agents a way to verify their results in a feedback loop as short as possible, along with few instructions for the important guardrails, and a way to organise efficiently (a ledger, a GitHub repository with issues, Linear, Jira, whatever.) With that, you can tend to 3-5 parallel sessions to avoid waiting, and actually get stuff done fast.
To me the ones doing scheduling, team coordination and chasing people down for answers to promises made are probably the safest. The more you deal with humans in meatspace the safer from AI you are.
So what happens when everyone is building the "same" app because they ask the "same" question to AI, without actually using their understanding and creativity to be "different" ?
My hunch is that software is pretty much going to end the way we think of it today.
Hardware, too, but that will take longer. If you think big enough, the idea of waiting for a new TV or phone or gaming system when you can just have your agent take your description, spec and order the hardware and have it fabbed for you in days ... is kind of quaint.
The broad societal disruption will be pretty devastating. But looking at software as the microcosm really lets you dream (/nightmare) big.
To me the question is not will we be able to get anything we want how we want it, but more who will have money to buy this?
The real question for executives is: engineer + AI or PM + AI?
I certainly believe that decent engineers do already have a good degree of knowledge about product topics (it’s like a must nowadays). PM/POs don’t really know enough on the tech side, so they would be clueless even with claude.
This seems like a very doomer take. Hard to take this assessment seriously with absolutely no data or numbers to back this up. Especially since devs certainly don’t seem like they are going to be “cooked”.
Execution of an idea is not the only function of a PM. There is product development, customer feedback, outreach, go to market, customer success. I don't see how AI is wholesale replacing customer outreach & go to market but I could be wrong.
Product Managers figure out what people want. Using AI they can easily transition to become founders. The decision Big Tech has to make when hiring one is: Would I want to compete against this person.
That's what my last company told themselves. Five different simple client/server applications in its suite, five different architectures with different language/framework/storage choices. At one point, my senior engineer spent 6 months writing his own Java threading library because he thought he was smarter than Java devs.
I was hired as a TPdM. I unabashedly wrestled technology decisions away from them because these "senior" and "principal" engineers obviously showed they could not be trusted.
So PMs aren’t safe because you need to…understand customer pain points?
Buddy, that skill is what PMs that are any good at their job spent a lifetime perfecting.
Rather, PMs aren’t safe as managers could just ask the agent to find the problem worth solving and then go solve it. But that takes us to a rather different world.
There has been writing on the wall for non-technical PdMs for years now, especially at small, mid-size companies. They lack the technical chops to eliminate friction for engineering teams, and I've witnessed way too many examples where they don't technically align their products to the greater enterprise ecosystem.
We're in interesting times. Many large, non-tech companies digitized ~30 years ago by building instead of buying. We're at a stage where many are now forced to pay off the technical debt accumulated since then, and many realize that it wasn't just engineering that's responsible for the debt, it was non-technical PMs, too. Many PdM job descriptions are now looking for people from engineering backgrounds to fill those roles to avoid this in the future.
None of the articles, that purport to know who is safe and who isn't, should be taken at face value. We literally have no idea how anything is going to shake out, given how fast everything is changing.
The "job" that will still continue to exist will be the one who has to defend a company from getting their data breached against swarms of AI agents at >20,000 TPS.
Someone will have to be the technical bodyguard that needs to thwart an attempted breach as quick as possible without causing an outage and that is the difference between a $45,000 breach vs a $450M breach. Enough to bankrupt a startup.
Using something like Cloudflare won't be enough this time.
Exactly because professionals are often artists, who are creative instead of taking the same shot as everyone else. Insta is full of copy cats, same photo different "person". Its boring because of that, seen one seem em all.
IMO anyone who feels safe from what is going to happen (already happening, really) in the labor market over the next decade is ignorant or is heavily in denial.
I'm not saying we're surely all doomed here, but there are going to be a lot of changes that none of us can imagine every aspect of, and they are going to happen at a surprising speed, and there are a vast number of cascading impacts that people aren't thinking much about.
eg. if you're in the trades, yes you're safer than a knowledge worker for now, but if all the knowledge workers get wiped out you're both going to have vastly decreased demand for your work (far less individuals will have the disposable income to hire you, companies will have less need for facilities to house employees, so you'll also see a retrenchment on the corporate side) and you'll start getting a labor supply glut as the knowledge workers all desperately attempt to retrain for the 'safe' jobs. And this is just the short term problem, without getting into the robotics threat.
If we don't get ahead of this problem politically (spoiler alert: we won't, at least in the US, it is unclear if we even have a functioning democracy), things are going to get very messy for virtually everyone who relies on labor income.
The trades also have the difficulty of pricing themselves out of the market, while AI can come in and help normal folks to do all that stuff themselves.
Hopefully all these unpredictable side effects include bringing down the cost of living.
> Hopefully all these unpredictable side effects include bringing down the cost of living.
Hopefully, but I think it is going to be difficult for that to happen in a positive way because AI is likely to suppress wages for most (for the reasons you mentioned) but our entire current economic system is engineered to never be deflationary.
So it is hard to imagine a situation in which AI causes us to all earn less but still be equally or better off without some pretty fundamental changes to our economic system that will be hard to get consensus on in our current political environment.
Thank you for saying this. Business strategy will always ruthlessly optimize for efficiency. And this means doing more (or same) with less humans. Anyone not seeing this happening already isn't seeing the news.
60k fired from Microsoft
Today Cisco announcing 1 engineer using AI to try to replace/improve 98% of Cisco's code
Goog just lost $475B in value
These are not random disparate events. This is the industry changing to business as usual with less humans.
Note that people laid off are not able to find jobs for more than 9 months. This is a very serious situation.
The Product Manager profession has been made completely redundant in AI powered org. Everybody can look at numbers, plan features, and push for them. All the AI redundancies are focused on eliminating middle management.
There is absolutely no place for PM. I say this as someone who has been in PM leadership and built product orgs several times over the last 20 years.
idk how common this is in the industry though, but since the introduction of AI my role has expanded so much beyond what is just writing code, i'm now a devops/product/dev person and expected to do all of that at once for the same pay as before
i actually enjoy it myself cause you feel less like a ticket pusher
As long as humans are making decisions, human interaction remains important. From this perspective, PMs are still in a better position than engineers
For the rest of us, that's not a problem.
What should we call these intermediate deterministic artifacts? Code!
Yeah we'll still need to know a sequence to flip bits in the right order, but it isn't code you'll recognize.
I think you can eventually set up an AI system that handles a lot of the interactions and conversations people have.
> I grew up with the distinction of "tech guy" needs a "product guy". I am not dismissing it, I am just saying that the world is entering into a phase where both roles are eventually melting into a single function.
The whole reason why Engineering exists in organizations is to take Product's ideas and turn them into reality. When Product can realize its own ideas (through AI) then Engineering is no longer necessary as a political sub-organization. Making it happen requires engineers to learn Product skills and adopt a "Product" job title. We will likely call people with both skillsets Product Engineers. People who can't adapt to learning both skillsets are at risk of losing their jobs.
Gen AI has changed every single part of classic software development, and the impact hasn't even remotely been felt. And zero jobs are "safe" in that traditional way that we saw them.
It's actually staggering to think of how the teams I've led developed software five years ago, ten years ago, fifteen years ago and twenty years ago. Easily 95% of that effort has been obliterated (I'm estimating on the low side, more like 99.5%). Classic wisdom like "the specification needs to be 100% up front because once we implement it stratifies" are relics of the past.
Everything has changed. I can't fully express how software development will work, but it certainly won't involve someone whose job is "human interaction".
For security or applying business logic (and how it interfaces with programming or database logic), you get less and less assurance by just looking at it.
Citation needed. Many of my users don't even know what correct would be.
> nobody is going to try building a building only to check if it's structurally sound because that would be absurd.
Well, that otoh can be simulated quite easily.
Neither of these statements are true if your team is building anything remotely complex with real customers. You can’t quickly verify if your decisions will survive the test of time and past mistakes can’t always be easily fixed. LLMs trivialize the easy parts of software development and enable faster enshittification of the projects they’re used on, but that’s about it.
> On the other hand nobody is going to try building a building only to check if it's structurally sound because that would be absurd.
Structural engineering isn't some mystical science. It's a set of mathematical truths, calculations, SOPs, and so on. If a human can do it, of course a machine can do it, and I'm obviously not saying "run it through the stochastic parrot and out comes a corporate headquarter building", but it's a part of the process where we're rapidly going to get there. And of course, the disruption in that field is already underway. For instance Stru AI. Yes, an engineer is still involved, but the premise is that your single engineer now replaces 10 engineers, and so on.
Gen AI isn't going to be plunging toilets or changing bedpans...yet...but it disrupts almost every role in the "white collar" world. That doesn't mean it eliminates every job, but it completely upends the norms.
AI can now push new features at speed, but somebody still needs to steer those features into a desirable product. Customers will ask AI for a faster horse.
This is why AI companies aren't spinning out new companies left and right. They can obviously push huge volumes of code, but can they effectively manage the resulting products?
Huge volumes of buggy code. The ChatGPT web interface should be an absolute embarrassment. Tap the input box, scrolls to top. Submit input, scrolls to top. 1+ second delay from keypress to input. Scrolling locks up. It’s been an absolute buggy mess for ages, the specific bugs change day to day. I suppose this is what happens when people think the LLMs are good enough to let loose without supervision.
> Code spitters are cooked, let's be honest here and accept the fate so everyone can move on.
I'm guessing "code spitters" here means people who consider their job to be manually writing code based on requirements produced by others.
Personally, I think the role of "developer" has already expanded way beyond that. If you're a technology expert who understands what is and is not possible to build, how it can be built, how to prompt that thing into existence, verify that the implementation is good, then get that implementation safely into production, you're not "cooked" at all.
Is this just domain dependent?
Code quality improves month by month, and by now I rarely have complaints in PR reviews, even if I read through them manually, line by line. For the most part, it's how I would write it myself, only that I would have been slower, and not as motivated to test it so thoroughly.
The key is IMHO giving the agents a way to verify their results in a feedback loop as short as possible, along with few instructions for the important guardrails, and a way to organise efficiently (a ledger, a GitHub repository with issues, Linear, Jira, whatever.) With that, you can tend to 3-5 parallel sessions to avoid waiting, and actually get stuff done fast.
The "I just manage the Gantt chart" folks are screwed, I think.
The ones that are good at translating between teams can be a huge force multiplier.
My hunch is that software is pretty much going to end the way we think of it today.
Hardware, too, but that will take longer. If you think big enough, the idea of waiting for a new TV or phone or gaming system when you can just have your agent take your description, spec and order the hardware and have it fabbed for you in days ... is kind of quaint.
The broad societal disruption will be pretty devastating. But looking at software as the microcosm really lets you dream (/nightmare) big.
To me the question is not will we be able to get anything we want how we want it, but more who will have money to buy this?
I certainly believe that decent engineers do already have a good degree of knowledge about product topics (it’s like a must nowadays). PM/POs don’t really know enough on the tech side, so they would be clueless even with claude.
Many developers don't want to think about these at all (and I say this as a former-developer-current-PM).
I was hired as a TPdM. I unabashedly wrestled technology decisions away from them because these "senior" and "principal" engineers obviously showed they could not be trusted.
Buddy, that skill is what PMs that are any good at their job spent a lifetime perfecting.
Rather, PMs aren’t safe as managers could just ask the agent to find the problem worth solving and then go solve it. But that takes us to a rather different world.
We're in interesting times. Many large, non-tech companies digitized ~30 years ago by building instead of buying. We're at a stage where many are now forced to pay off the technical debt accumulated since then, and many realize that it wasn't just engineering that's responsible for the debt, it was non-technical PMs, too. Many PdM job descriptions are now looking for people from engineering backgrounds to fill those roles to avoid this in the future.
And you are 100%, this was thoughts wandering after a long late night discussion with friends who work for a CRM company.
Not backed by data, not a prediction, just subjective take at 1am.
And yes, thinking architecture, tradeoffs, scalability and safety isn't going anywhere.
Spending all day writing code as proof of work is out.
Someone will have to be the technical bodyguard that needs to thwart an attempted breach as quick as possible without causing an outage and that is the difference between a $45,000 breach vs a $450M breach. Enough to bankrupt a startup.
Using something like Cloudflare won't be enough this time.
I'm not saying we're surely all doomed here, but there are going to be a lot of changes that none of us can imagine every aspect of, and they are going to happen at a surprising speed, and there are a vast number of cascading impacts that people aren't thinking much about.
eg. if you're in the trades, yes you're safer than a knowledge worker for now, but if all the knowledge workers get wiped out you're both going to have vastly decreased demand for your work (far less individuals will have the disposable income to hire you, companies will have less need for facilities to house employees, so you'll also see a retrenchment on the corporate side) and you'll start getting a labor supply glut as the knowledge workers all desperately attempt to retrain for the 'safe' jobs. And this is just the short term problem, without getting into the robotics threat.
If we don't get ahead of this problem politically (spoiler alert: we won't, at least in the US, it is unclear if we even have a functioning democracy), things are going to get very messy for virtually everyone who relies on labor income.
Hopefully all these unpredictable side effects include bringing down the cost of living.
Hopefully, but I think it is going to be difficult for that to happen in a positive way because AI is likely to suppress wages for most (for the reasons you mentioned) but our entire current economic system is engineered to never be deflationary.
So it is hard to imagine a situation in which AI causes us to all earn less but still be equally or better off without some pretty fundamental changes to our economic system that will be hard to get consensus on in our current political environment.
60k fired from Microsoft
Today Cisco announcing 1 engineer using AI to try to replace/improve 98% of Cisco's code
Goog just lost $475B in value
These are not random disparate events. This is the industry changing to business as usual with less humans.
Note that people laid off are not able to find jobs for more than 9 months. This is a very serious situation.
Not true
There is absolutely no place for PM. I say this as someone who has been in PM leadership and built product orgs several times over the last 20 years.
If you rely on AI to make those decisions, at least with today's capabilities, you will quickly see how sub optimal the results are.