Help My Job is Changing! A Complete Handbook to Product, Engineering, and Design Role Changes Driven by AI

The most common question I get is "how will my job change with AI?" I wrote this post to help. Think of this as a complete handbook for what to expect as AI brings changes for building roles in tech!

I think this is a really important post, so please share it with people you know who are worried about how AI is changing building roles. And before we go farther—YES I have a big scary FAQ at the end where all the scariest questions get addressed.

I could have written it as three separate posts, but I think the lines are starting to blur for engineering, product, and design, and we need to talk about it really honestly. With that in mind, dig in! Feel free to jump to your particular role or read all the way through—you’ll get interesting nuggets either way.

As long as this is, I made it short! There is so much more to say here, so I will probably write more on this as AI continues to shift all our roles. Let me know in the comments what you’d like me to dig into more!

The Evolution of Building Roles: How AI Is Redefining Product, Engineering, and Design

Picture a product meeting in the not-too-distant future: real-time AI analytics are summarizing millions of data points into actionable insights, code is auto-generated to fix known bottlenecks, and rapid design variations arrive with a single click—each one curated by algorithms that have already considered user feedback. None of this is science fiction anymore. Nor is it a sales demo video (ha!) Within the next two to three years, artificial intelligence will not merely refine how we work; it will thoroughly reinvent the roles and career paths of product managers, engineers, and designers.

From SaaS startups to massive healthcare enterprises, the move toward AI-driven product development is reshaping daily tasks, required skills, and how each discipline collaborates. In place of siloed responsibilities and linear career ladders, we’re heading toward overlapping roles, “human + AI” workflows, and a faster, more integrated product life cycle. This article unpacks that shift, drawing on the latest research and real-world examples to show how the quiet AI revolution is already reshaping our jobs—and how to position ourselves for success in this changing landscape.

1. A Rapidly Changing Landscape

For decades, product managers, engineers, and designers worked in well-defined silos: product managers focused on strategy and requirements, engineers wrote code, and designers polished the user experience. As AI spreads to every layer of product development, those boundaries are dissolving—revealing at least three powerful shifts.

Cognitive Offloading

AI handles much of the busywork once considered part and parcel of these roles. From poring over user data and automating routine market research, to generating boilerplate code and even crafting initial design variations, AI systems are taking on tasks that can be automated. This “cognitive offloading” frees human team members to concentrate on what truly adds value—creativity, strategic thinking, ethics, and problem-solving.

Skill Convergence

Because mundane tasks can be offloaded to AI, roles now bleed into each other. Product managers may code prototype features. Designers learn to interpret data analytics. Engineers help shape user experiences. It’s no longer bizarre to see a designer using AI to create multiple interface mockups in minutes, while a product manager fine-tunes data-driven marketing copy the same afternoon. Companies increasingly seek “AI-savvy” talent—people who can combine product sense, design sensibilities, and technical fluency.

Dynamic Career Paths

The convergence of responsibilities is creating hybrid titles like “AI Product Strategist,” “AI Ops Engineer,” and “Design Technologist.” Career growth is becoming less about how many years you’ve spent in a narrow specialty and more about the breadth of impact you can achieve with AI. Those who embrace AI-driven workflows—making bold decisions, collaborating fluidly across domains, and championing responsible innovation—will climb faster than ever.

2. How AI Is Transforming Product Management

Product management is already heavily data-driven, a trend AI will only amplify. According to a recent study, 92% of product managers foresee AI having a “huge impact” on their work. In other news, the sun rose today. In practice, AI tools are freeing PMs from routine data tasks, enabling deeper strategic input, and expanding ownership of critical cross-functional issues.

2.1 SaaS

In the SaaS world, product managers have historically wrestled with reams of product usage data, user feedback, and competitive analysis. AI changes that equation in two key ways. First, it automates the drudgery of sifting through endless analytics; an AI assistant can glean insights—such as new feature priorities or patterns of user churn—and then propose potential solutions. Second, the ubiquity of AI means that AI literacy becomes a required skill. Marty Cagan made the sort of obvious claim that the PM role, especially in SaaS, will inherently be an “AI PM” role going forward. I agree, but I think that’s not enough detail to be useful.

Anyway, because AI can also generate design mockups and marketing copy, the scope of a PM’s responsibility widens. Rather than delegating every small design request to a separate team, a savvy PM can prototype quickly. Paradoxically, this also raises ethical and governance challenges: if you’re using AI to generate user-facing features, how do you ensure fairness, accessibility, and privacy? In SaaS, a PM may become the chief guardian of responsible AI usage and data compliance—reviewing AI-driven features for bias or potential misuse before launch. Just hitting approve may generate headlines down the road that companies don’t want, but it’s going to happen.

2.2 Fintech

Fintech runs on large-scale data about transactions, credit histories, and risk models. As AI takes over tasks like real-time fraud detection or the personalization of loan offers, PMs in fintech find themselves at the intersection of innovation and compliance. They must ensure their AI-based features comply with regulations (e.g., anti-discrimination in lending) while delivering a seamless user experience. A product manager here might spend half the day working with data scientists to tweak risk assessment models, and the other half collaborating with legal teams to document how decisions remain fair and transparent. And the third half (ha!) desperately trying to write go-to-market strategy while waiting on the bank to get back with approval.

Over the next few years, PM roles are going to shift even harder to monitoring AI models, spotting anomalies in real-time, and testing new data-driven features. That means more rapid skill gain of domain knowledge, so that fintech experience that once took years of manual analytics to accumulate now might take a few months of intentional AI partnership. Simultaneously, PMs will face AI as the ultimate complexifier: boards will demand it, but it will present as an extra complex dimension across multiple lines of business—merging compliance, user trust, and technical feasibility into cohesive product strategies.

2.3 Consumer Tech

Consumer tech thrives on personalization, speed, and constant iteration. AI amplifies all three. Think hyper-personalized feeds, dynamic user interfaces that adapt to each individual, and voice/AR interfaces that rely on real-time machine learning. For the product manager, this creates a heightened level of accountability. When an AI-driven recommendation feed goes awry, user trust plummets—making ethical considerations and transparency major concerns. Always-on AI also heightens expectations for consumer-facing ships to be near-instant. If the AI discovers a bug and files it, how quickly is the PM held accountable for getting that fixed? Or what about an AI-filed feature request? It’s going to happen.

And continuous deployment cycles are accelerating. AI might not just propose bugs or file tickets. AI can propose new A/B tests or automatically roll out (and roll back) certain features. Product managers must orchestrate this environment, ensuring that “move fast” does not degenerate into “break things”—especially in areas of content moderation or potential bias. This shift can compress product life cycles, letting PMs take on significant, AI-driven features more quickly than in the past. If the tests win the outcomes might turbocharge careers—but winning PMs will need to skill up on AI metrics, user psychology, and data governance.

2.4 Healthcare

In healthcare, data is abundant—clinical notes, lab results, imaging scans—but deeply regulated and often sensitive. Product managers must already navigate HIPAA, FDA oversight, and a wide network of stakeholders (clinicians, patients, insurers). AI is enabling new possibilities—like apps that monitor vital signs or AI-driven diagnostic tools—but PMs still have to face human bureacracy. They still have to carefully validate their products through clinical trials and regulatory approvals. Again, AI arrives as a complexifier: how do you guard against algorithmic bias in a disease-detection feature, and how do you present AI recommendations to clinicians so they trust, but also verify, the results?

Many healthcare PMs will become “translators,” bridging data science teams and medical professionals. While the learning curve can be steep—absorbing both clinical vocabulary and AI fundamentals—those who succeed might (maybe) deliver enormous social impact. Moreover, because the release cycles in healthcare can be longer (due to the need for clinical validation), PMs in this space get good at juggling near-term improvements with multi-year AI product roadmaps. And by the way, I don’t think AI will magically shorten that cycle!

2.5 Hard Tech

Hard tech spans sectors like robotics, automotive, aerospace, and IoT devices. Product managers here face technical complexity and physical constraints, from materials and manufacturing to safety regulations and distribution logistics. AI helps accelerate R&D: product teams can run complex simulations before building physical prototypes, or use generative design tools to propose new parts optimized for weight or energy efficiency.

There tend to be less PMs here. For PMs that do work in hard tech, this means you become part-technical architect, part-supply-chain coordinator, ensuring that AI-driven design changes are actually manufacturable at scale and cost-effective. The role can also demand cross-company negotiation—deciding whether to build proprietary AI modules or partner with specialized vendors. The result is a far more integrated role than the traditional hardware PM of ten years ago. Those who can speak the language of AI, hardware engineering, and user experience simultaneously will be best positioned for growth, but it’s a big ask.

3. How AI Is Transforming Engineering

Gone are the days when entry-level engineers toiled exclusively on rote coding or repetitive QA tasks. AI coding assistants can now handle much of that. Across SaaS, Fintech, Consumer Tech, Healthcare, and Hard Tech, engineering roles are moving “upstream” into architecture, integration, and critical problem-solving.

3.1 SaaS

Engineers in SaaS are rapidly moving to Cursor. The result: improved productivity and a shift in what day-to-day work looks like. Engineers still write code, but (a varying amount) will be generated or suggested by AI. How much? Depends on your code base. Depends on the system rules in Cursor. And who are we kidding—Cursor has won here. Depends also on your proficiency in figuring out where it is efficient to write the code with AI. Their real challenge is how engineers can grow into a new skill set: becoming orchestrators and planning the system design so that AI outputs are correct and maintainable at large scale.

With these shifts, the bar for engineering quality rises. Yes, that should scare you a bit. Mediocre coding gets automated away. Engineers set themselves apart by devising novel solutions, debugging complex edge cases, or constructing entire data pipelines that feed AI algorithms. Because SaaS often requires rapid iteration, AI-augmented coding and testing will be expected to shorten release cycles, and engineers will have to figure how to operationalize that board expectation. Ultimately, a continuous deployment environment rewards engineers who can move fluidly from concept to production without lengthy handoffs.

3.2 Fintech

Fintech engineering deals with real-time fraud detection, risk modeling, and stringent compliance demands. AI can flag suspicious transactions in microseconds or automatically generate compliance checks within the code. Engineers in Fintech thus spend less time writing low-level validation scripts and more time interpreting the outputs of machine learning systems—and ensuring they integrate seamlessly with legacy financial infrastructures.

As a result, entry-level “operations” or “IT support” roles might shrink, replaced by “AI Ops Engineers” or “MLOps” specialists who build and monitor the automated systems that keep financial platforms secure and compliant. Senior engineers handle risk at a broader scale—ensuring the models are explainable, auditable, and do not inadvertently discriminate. This combination of deep technical skill and regulatory awareness is relatively rare; but for engineers who can develop it, that skill combination can pave the way to fast-tracked leadership roles. And yes, I’m aware engineers may not want to be leaders (ha).

3.3 Consumer Tech

In consumer tech, massive user bases and quick iteration cycles make AI indispensable. AI-driven testing, auto-generated unit tests, and real-time monitoring can detect issues before they spread to millions of users. Engineers focus on system design, building personalization features, or orchestrating complex data flows that drive content feeds and recommendation engines.

Front-end engineers are no exception. As AI moves to client-side personalization and on-device inference, front-end specialists will be expected to integrate smaller ML models directly into apps—and that’s going to require a solid LLM engineering background. Engineers will need to get into edge computing, on-prem models, stable inference across a range of devices and a whole mess of model quality management.

3.4 Healthcare

Healthcare engineering is safety-critical: an erroneous AI output can endanger patient lives. This drives an emphasis on validation, privacy, and robust testing. Engineers might rely on AI to simulate thousands of clinical scenarios, reducing the need for endless physical prototypes. Yet human oversight remains paramount. Every proposed AI diagnosis feature undergoes stringent real-world trials, with engineers and clinicians collaborating to confirm safety and effectiveness.

Engineers in this field will also have to double down on privacy safeguards—integrating encryption, federated learning approaches, and strict access controls. Despite these constraints, the potential upside is immense. The ability to ship an AI-enabled app that improves patient outcomes is a powerful motivator, and that may attract engineers into the space who care about the outcome they generate in the world—not just the paycheck. And because the domain is complex, engineers with experience in health data formats (like HL7/FHIR) and AI tools will command stronger compensation.

3.5 Hard Tech

Hardware-centric engineers—whether in robotics, automotive, or manufacturing—are increasingly turning to AI-driven simulation and generative design. Instead of painstakingly modeling each iteration, they define constraints and let AI propose multiple designs to test in virtual environments. Engineers then refine the designs that look most promising. Hey—did you notice that’s a lot like what PMs do in hard tech? Yep, the lines are blurring!

These new workflows shrink development cycles, reduce costs, and encourage bolder experimentation. However, they also demand deeper systems thinking. A robotics engineer, for instance, must integrate AI-based vision systems, servo controls, and mechanical design in one cohesive system. And because hardware products often continue to evolve post-shipment (through firmware updates or on-device AI models), engineers remain involved in continuous improvement long after the product hits the market. The role can start to feel more like software—with version updates, real-time telemetry, and user analytics—further blurring boundaries between hardware and software expertise. ASML ships an engineer along with its etching machines for a reason!

4. How AI Is Transforming Design

Designers are sometimes the first to worry that AI will “replace” creative work. Yet what’s emerging is far more collaborative than annihilative: AI becomes a creative partner, automating production-level tasks and liberating designers to focus on curation, strategy, and the profound human aspects of design. Taste still wins, guys. And I know some of you in the comments will tell me that AI can automate taste, and maybe eventually it will, but over a reasonable career horizon I see a strong alpha in taste.

There’s an extra twist on top of taste—designers across the board are going to need to lead the way on design taste for both humans and agents! How do you design a system that is ground-up workable for both autonomous AI agents and also humans? Designers will be expected to know the answer by 2026.

4.1 SaaS

In SaaS, product design often moves at a breakneck pace, with features rolling out continuously. AI amplifies that speed. Designers can generate multiple layout variants, style options, or entire UI components from a single prompt. This helps them iterate rapidly, testing different concepts with minimal manual overhead. They then use their human intuition to refine the best AI outputs. The result is better exploration of the design space, faster.

In parallel, AI contributes to adaptive UX systems. SaaS products might adjust interfaces based on real-time analytics: if a user is struggling, the product’s UI reconfigures to help. Designers oversee not just static mockups but living systems. Their main challenge becomes curating a designed experience that serves consumers as it lives, breathes, and evolves with changing user behavior. If this sounds close to PMs—the lines are blurring! See the theme?

4.2 Fintech

Design in fintech is all about trust. If users don’t feel comfortable, no amount of fancy AI will win them over. AI helps by parsing complicated financial data into digestible visuals or personalizing app flows for different risk tolerances. Yet designers must ensure that any AI-driven suggestions remain comprehensible—why was this loan denied? why did this transaction flag as suspicious? Designers therefore become guardians of transparency.

They also manage constraints from regulators, sometimes needing to embed legally required disclaimers or follow strict style guidelines for certain disclosures. That part isn’t new, and AI can automate parts of the boring bits—say, by drafting explanatory text—but the final editorial oversight belongs to the human designer. And as personal finance tools branch out into voice or chatbot interfaces, designers skilled in conversation design and complex data visualization are going to become very valuable.

4.3 Consumer Tech

In consumer-facing apps, AI accelerates creative exploration. From ephemeral AR filters to user-generated content, design often merges brand aesthetics with user spontaneity. Designers can use generative AI to create visual assets at scale—like a dozen new sticker packs or an array of styling for profile pages—then refine them to maintain a sense of brand identity. Designers also have to grapple with personal AI agents as a factor now—how do you change the visual identity of the site when a user’s ChatGPT 5 or 6 companion is browsing for them?

And the line keeps blurring with PMs here: designers are going to be expected to be accountable to business outcomes driven by AI, and they’re going to be asked to lead with strong visual identity without compromising the AI capabilities of the systems they’re building. Sometimes that may look like strong Design-PM pairings, and sometimes one or the other will do both roles.

4.4 Healthcare

In healthcare, design decisions can influence patient safety as much as aesthetics. AI can crunch feedback from thousands of patient interactions to surface usability issues—perhaps a confusing button layout or a poorly worded alert. Designers must then adjust to ensure clarity and empathy. They also have to respect clinicians’ workflows, providing just enough AI-driven guidance without overshadowing professional judgment. This is really really hard problem, and we’re still figuring out how to guide and optimize medical provider decisioning in ways that increase true positive diagnosis rate, especially for medical edge cases.

Moreover, healthcare design teams follow strict validation protocols, sometimes running simulations with clinicians. A single interface tweak can require sign-off from medical, legal, and regulatory stakeholders. Although that pace is slower, the resulting impact can be profound. A well-designed, AI-assisted app can detect early signs of chronic disease, streamline hospital triage, or help patients manage medication. Designers who master this balance of human empathy, clinical accuracy, and AI-driven insights can become influential leaders in the healthtech domain.

4.5 Hard Tech

Hard tech designers frequently tackle industrial design, ergonomics, and the interplay of physical and digital elements. AI-driven “generative design” tools can suggest novel shapes, materials, or configurations. We now have text-to-CAD for quick design iterations! The result can be more innovative aesthetics—lighter structures, eco-friendly materials, or unexpected geometric forms—without sacrificing structural integrity. Designers then fine-tune AI outputs, preserving brand identity and addressing manufacturing constraints. Yes this gets blurry and somewhat close to engineering roles in hard tech. An engineer might check or review a designer’s part design for structural integrity now, whereas before the engineer would have to spend days in CAD drafting the concept from the designer.

Hard tech also opens the door to mass customization. AI can quickly generate designs tailored to each user’s measurements or preferences (e.g., ergonomically fitted wearables, custom car interiors). Designers architect the parametric rules that let AI vary product dimensions while remaining faithful to safety and brand guidelines. This demands not just visual creativity but also logical, system-oriented thinking. Designers who win in this space will not be afraid of going into materials science so they can accelerate value through the iteration cycle.

5. Cross-Functional Trends and Converging Skill Sets

Though each role and industry has its unique pressures, the AI revolution is producing broad, cross-functional patterns:

Blurred Boundaries Are Here to Stay: Surprise! Product managers, engineers, and designers increasingly share workflows. A PM might tweak a design prototype via an AI tool; an engineer might incorporate user feedback directly into code; a designer might analyze real-time usage metrics. AI acts as a collaborative layer, making each role more capable of straying outside its old silo.

Hybrid Roles Are Real: From “AI Product Strategist” to “Design Engineer,” new hybrid titles reflect the intersection of creativity, technical knowledge, and product management. Companies looking to hire now often list a mixture of soft and hard skills—communication, domain knowledge, AI literacy—in a single job description Meeting those expectations is an unclear bar, and often hiring managers don’t exactly know what they want.

The Future is Flat: With AI handling routine tasks, fewer layers of bureaucracy and fewer specialized sub-teams are required. Cross-functional “pods” may move quickly, each member wearing multiple hats. Teams can get reassigned fast. Teams may move with dedicated AI agents that hold context.

Promoted on Impact: Instead of a traditional ladder, especially early on—“spend X years as a junior, then Y years as a mid-level”—career advancement increasingly will depend on tangible contributions. Launch a groundbreaking AI-driven feature that boosts key metrics or solves a major pain point, and the promotion might come sooner than you’d expect. Many organizations now reward the ability to integrate AI effectively into daily work over raw tenure. The other side of the coin is that if you don’t have those outcomes to show, you’re going to get booted fast.

6. Innovating Your Future: Beyond the Status Quo

A Vision of Augmented Roles

Envision a near-future workplace where every professional is augmented by AI. Product managers derive instant, data-driven insights on market shifts; engineers collaborate with AI coding assistants for robust, secure code; designers generate dozens of potential solutions at the push of a button, refining them to produce experiences that resonate deeply with users. In this scenario, human creativity and AI computation complement each other rather than compete.

Rather than eliminating roles, AI is shifting the focus of each role to higher-value tasks: forging cross-disciplinary collaboration, championing ethical standards, and orchestrating innovations that genuinely improve people’s lives. The quiet revolution already underway will only intensify over the next two or three ye

The Scary Question FAQ Section

This section is dedicated to openly asking and answering the scariest questions I get about these three roles as plainly as I can.

1. What will happen to junior roles?

I think we will have fewer of them. If a strong junior PM can use GPT 5 or 6 and do the work of six juniors, well we need only one junior! This trend is accelerating a broader trend in tech around hiring only experienced senior staff. AI is going to accelerate that. Junior role winners here will be showing extraordinary impact before getting into the role (I know!)

2. Are wages going to be impacted?

Yes, I think so. But in weird ways. I think we will see a strong split in roles between AI-enabled and non-AI roles. We already see this in compensation splits for roles associated with AI today. Eventually I’m expecting a roughly 10x pay boost for star performers powered by AI in any of these three job families. And there will be increased downward compensation pressure down the chain as others with lower AI skills battle for roles.

3. What about senior roles?

Don’t sit on your achievements! You need to be investing stat in a mid-career transition to AI, and focus especially on showing outcomes. Senior roles are better positioned than junior roles in the bright new AI future, but it’s still going to be an outcome-driven world, so obsess over what you can ship.

4. Will there be fewer jobs?

I don’t know. We will probably know by 2026. I think my current bet is that total tech open roles will very slowly decline to roughly 75% of 2019 levels over the next 2-3 years. The slow decline in postings will mean high competitive pressure (discussed above). I think there will also be lots of new roles opening up within that overall mix, and people with strong AI skills are positioned to make bank. On a given team, I think this looks like delayed hiring, only hiring one person when you hired two before, etc.

5. How will promotions work?

I think leadership transitions are going to be broken for awhile. It was always hard. But now a leader has to manage AI agents, probably. And also people. And the number of teams and team sizes are both shrinking, so there are fewer management opportunities. In particular, I think Directors are an endangered species. Middle management is about information repurposing, and AI is very good at that. But that leads to a question we can’t answer yet: how do you get ready for senior leadership roles without junior leadership roles? We don’t know yet.

So there you go! That’s my take on some of the hardest questions I get on AI and jobs. What did I miss? What do you think will happen in product, engineering, and design? Let me know in the comments!

It really is going to feel like an atomic era change for jobs