Meta’s AI ambition and the industry impact
At the beginning of the year, Mark Zuckerberg revealed on a podcast that engineers at Meta are developing AI capable of performing as mid-level coders; AI that, throughout 2025 and beyond, could ultimately replace human engineers. Weeks later, Meta began swiftly cutting its staff in favor of more AI/machine learning engineers. Since then, several technology companies have downsized, citing AI as a reason (e.g., Salesforce, Hewlett Packard, Workday, etc.).
The rapid advancements in AI are understandably generating a lot of discussion. Just take one scroll through Reddit (I’m admittedly on this a lot…), LinkedIn, or any open forum, and you’ll find an abundance of opinions on the future of engineering and the tech industry.
Crowd-sourcing POVs on the topic
This inflection point has understandably come up a lot in my line of work recruiting engineering and product executives. I often bring up this topic to many of them to gather their points of view. I’m not an engineer, nor do I pretend to be (pre-med Bioscience major, to law school, to lawyer, to recruiting ... story for another day), but I am hungry to learn and create a framework on what skills and traits will be needed for the future engineer.
Here’s what I asked some product/engineering leaders in my network and what they had to say:
CAN AND WILL AI "REPLACE" MID-LEVEL CODERS?
“It’s overzealous and there’s no way Meta will cut their engineers this year for AI. AI is just another tool; people who know how to use it will be effective, and those who don't will be left behind.
It's similar to Assembly → C → Java → Python. Each represented a significant order of magnitude and increase in efficiency until everybody just learned the next tool.”
- CTO at a VC-backed edtech platform
“Jury’s still out, but I expect AI could ‘replace’ only in the sense that, by the end of this year, the profile of the engineer as we know it will be more of senior AI architects. The profile will be people who know how to architect systems beautifully as well as brilliant young talent out of school. So, those who don’t embrace or leverage this talent or AI will be in trouble.”
- Chief Product & Technology Officer at an intelligent conversational platform for P&C insurance
“AI won’t take your job. But someone using AI will.”
- CTO at a VC-backed retained search firm powered by technology and network effects
WHAT CAN THE FUTURE STATE OF ROLES AND RESPONSIBILITIES BETWEEN AN ENGINEER AND AI LOOK LIKE?
“While AI can accelerate or even complete coding tasks, it wouldn’t overhaul the entire software development lifecycle. We will likely just need to see the redefinition of roles within the industry, with one of the most significant shifts from mid-level engineers that code to more problem-solving roles.”
- CTO at a PE-acquired multi-billion dollar fintech company
WHAT ARE THE KEY TRAITS TO LOOK FOR IN THE NEXT GENERATION OF ENGINEERS?
“We need to be thinking about the talent coming out of the younger generation and students. There needs to be that innate brilliance. Look for those who are still bullish on software engineering and those that deeply understand the full stack on a foundational level. It shouldn’t be a 1-time trial with AI. We all need to keep investing and practicing in this jagged-edge frontier.”
- Chief Product & Technology Officer at an intelligent conversational platform for P&C insurance
HOW SHOULD YOU THINK ABOUT HIRING ENGINEERING/TECH TALENT IN LIGHT OF ALL THIS?
“I think the entry-level engineer probably needs to be retooled to become AI engineers versus just software engineers. Or maybe it’s a combination of both. There’s also hardware, which might not be fully integrated with AI, but there’s still an opportunity there for some significant reskilling across the board.”
- Global Head of Product at a top 5 biotechnology company
“If anything, there’s more likelihood of replacing entry-level, green engineers that often have no work experience and are typically more textbook. It’s much harder to replace a mid-level because they do so much more than just code. However, if you disrupt that entry-to-mid-level pipeline, then at some point down the line, you might not ever be able to find any qualified senior engineer.”
- Director of Engineering at a FAANG
“I’d look for talent who know how to truly own something despite AI. I still believe code is craft, and when I hire someone, I want them to own delivering something to me. Whether it’s coding or guiding LLMs, they just have to show the curiosity and ownership to push things through. I also look for a demonstrated interest in these new AI tools. If there’s resistance to that, it’s a red flag.”
- Chief Product & Technology Officer at an intelligent conversational platform for P&C insurance
TLDR
The points of view are not surprising. It's true that AI's capabilities are expanding. At the same time, history has shown us that technological progress doesn't lead to complete obsolescence but rather to evolution.
AI enables and equips engineers with tools that help them solve harder problems, faster. However, it’s still too early to predict if AI can “replace” engineers outright. We’re far and away from any endgame on this, if there even is one.
Engineering Focuses Before, Now, and… In the Future?
To help dive deeper, we zoomed out and took a broader look at how engineers’ allocation of time is evolving. (These are rough percentages and can vary depending on each team size and shape; the point is that the slices in the pie are changing in size and scope.)
Historically, almost half of a mid-level engineer’s workload could be easily dedicated to hands-on coding and development. Other important responsibilities such as system design/architecture, project management, maintenance, testing & QA made up the remainder.
But all of this is changing fast. Very fast. Now, AI is speeding up every slice of the pie — and that makes room for other slices like "Customer Insights" and things higher up in the value chain, e.g., deeper problem-solving and system design, cross-functional collaboration. While coding remains important, mid-level engineers at companies are now spending more time focusing on leveraging AI tools and making strategic implementation decisions rather than just writing code.
Just one example: JPMorgan Chase recently reported that implementing an AI coding assistant increased their software engineers' efficiency by 10% to 20%. This boost allowed engineers to allocate more time to high-value projects, particularly in artificial intelligence and data.
How to think about engineering talent moving forward
Returning to one of my initial questions (What are the key traits to look for in the next generation of engineers?), I decided to ask AI for its advice. Why not? 🤔 🤫 The inputs/prompts were built using:
- What I learned through conversations with tech leaders and from going deeper in the rabbit hole
- A few articles and other opinions I found online about what traits would be valuable for future engineering talent to have
- A few iterations of different prompts into the LLMs (Gemini and OpenAI)
In the end, the common theme across all the outputs was the notion of an engineer needing to fully embrace and possess early adopter traits in their DNA.
I immediately thought of the Product Adoption Curve (see below). There’s too much obvious logic to this conclusion: In a world of AI, engineers will have to be early adopters as things progress at breakneck speed.
This brings us to the next logical question. What ARE the traits of an early adopter?
Below is a simple table. Of course, mashed up via AI prompts and results.
The forward-thinking Product and Engineering Leaders I work with already prioritize these traits when building their teams. Meanwhile, some still remain fixated on whether a new hire knows the ins and outs of their specific technology stack. If anything, I foresee a future where the market becomes more uniform in its approach toward identifying and vetting these traits to complement varying proficiency levels of functional skills needed in the job. Naturally, these qualities can be just as crucial and transferable when hiring for other functions, including GTM, finance, operations, and people teams.
Again, I want to reiterate that great quote from our VP AI/ML (Oh, he created the Recommendation Engine for Spotify, by the way) that feels spot-on about this:
“Engineering skill levels generally range from zero knowledge → functional with oversight → autonomous → expert. AI has dramatically reduced the gap between ‘zero knowledge’ and ‘autonomously productive’ in many domains by providing in-time learning and real-world solutions."
These early adopter traits have always been around; there’s just more of a need now to have them in the spotlight and as a core focus to future-proof winning technology teams.
Engineering in the future?
So what happens next? How will the pie chart of engineering roles and responsibilities look like as AI takes a larger share of the work?
The reality isn’t as binary as "AI replaces engineers" or "does not”.
The engineers of the future will shift from coders to curators, architects, and strategists. The future engineer won’t just write code, but will design, guide, and govern the intelligence that does. How exciting is this?!
I wonder if it’s too late for me to switch careers again? After all, I am an Early Adopter with many things in my personal life💡.
