It's hard to be good at your tech job right now

I’m writing this post because I want to say out loud what a lot of people are feeling: Working in tech during the AI transition is tough as hell.

There are pressures from all sides. Here are just a few of them:

• Layoffs across the industry, especially reductions in management and support roles, as executives rush to replace human productivity with AI.
• Mandates to learn and adopt new tools and processes incredibly quickly.
• Shifting job requirements - like asking non-technical designers to become fluent with code-based workflows, or turning software engineering into an LLM-prompt-and-review lifecycle.
• Obligations to accelerate work, do more with less, and increase your output.
• A constantly changing market, where new breakthroughs and best-practice innovations launch nearly every week.
• An influencer-led hype cycle pushing AI as The Way, along with a subtle or overt denigration of traditional ways of working.
• A blurring of roles where non-experts can cosplay as experts because they created something with AI that looks credible on the surface.
• A structural reset in how people build things, what they build, and the time and effort it takes to complete each step of the process.
• Threats that entire job categories will be wiped out by AI advancements.

All of this adds up to a feeling of persistent intensity, and a sense of falling behind even when you’re working at max capacity. You have to rapidly adapt how you operate and how you spend your time.

Also: your job might no longer offer what attracted you to this line of work. If you loved to write code meticulously by hand – instead of asking a chatbot to write the code for you – you’re probably not super stoked about this movement.

And that’s not to mention the increasing public apprehension and negative impacts of AI on communities, which cast a looming shadow over the work whenever you have a rare moment to stop and think about it.

To top it off, the fragility of the job market leaves employees feeling more powerless to push back or share concerns openly.

Perhaps the most stressful side effect is the erosion of trust and confidence due to generative AI. It’s getting difficult to tell if something is real or fake, well-reasoned based on actual facts, or utter nonsense masquerading as truth. The worst case scenario is all of those at once. A truthy-looking document that’s half-accurate and half-bullshit is a dangerous document, because you need to look extra closely to spot the bullshit, and you might be accidentally lulled into believing it was all intentional.

Lately I’ve noticed my kids asking, “is this AI?” more often, and they’re increasingly skeptical of the information being presented to them. I find the same skepticism invading my work brain too. Did my coworker really think this through? Or did Claude Sonnet just half-ass a proposal based on some Granola notes?

We're facing an unstoppable onslaught of Brandolini's Law problems.

My kids are old enough to remember the world before AI existed, and maybe they’ll reflect on it like a before/after divide – similar to my memories of getting lost before we had always-on mapping devices in our pockets. It was scary, but we were unburdened by the shackles of constant ambient connectivity.

We were lost, but we were free to get lost.

The AI before-times will become similarly quaint. "Hey, remember when all information was authored by people?"

I know this post sounds pessimistic, which might be odd coming from a person who works at an AI-oriented company. That’s exactly the problem: tech workers are experiencing these internal contradictions every day. Rejecting AI makes you obsolete, but embracing it comes with a laundry list of hard compromises.

AI is not inherently bad, just like any technology is not inherently bad until humans decide to do foolish things with it. AI can be transformative when applied wisely, making it possible to solve previously impossible problems. That kind of sea change is always worth exploring with curiosity.

Unfortunately, most of the tech industry doesn't take time for doing things wisely. It's structured to chase one gold rush after another, and companies are usually willing to go scorched Earth to win, without much consideration for what breaks along the way.

So there's a vacuum of principled thinking, which leaves an opportunity for principled thinkers to fill the void. The AI movement is not particularly human-centered at the moment, and no one is presenting a counterculture argument for how it could be designed differently. Governments have little incentive or sufficient expertise to regulate anything. Frontier AI labs have all the control, but they're stuck between advancing the technology cautiously vs. building unfathomably valuable growth engines around it.

Let's remember that tech movements are not inevitable because some hype bros said so. Just look at the Metaverse or NFTs. It wasn't so long ago that web3 was going great. Tech only becomes inevitable when we throw up our hands, recede into futility, and allow the hype growth engines to roar unchecked.

Given the circumstances, the only way out is through. This has happened before: we pushed for web standards and accessibility, we pushed for health and safety in social platforms, we figured out the mobile revolution. Of course those things remain deeply flawed, but we made progress by speaking up and doing the work. This is a moment that requires the same. Our jobs are changing and so is everything else, but what hasn't changed is our responsibility to build with care and consideration for people first. It's hard stuff. But that's the job.

This is my second post in a series on how design is changing with AI. Part 1 is here. Stay tuned for part 3.