Big Tech, AI Everywhere, and Fewer Engineers: A Laid-Off Engineer’s Reality Check

Hello, world.After 25 years in the tech industry, I recently found myself laid off from Big Tech, an experience that has become increasingly common over the last couple of years. With some unexpected time on my hands, I’ve been doing what engineers do best: thinking deeply about systems, incentives, and failure modes.This time, the system in question is AI and more specifically, what the current AI boom means for software engineers.The AI Gold Rush Nobody Asked ForIf you’ve worked in Big Tech recently, you’ve seen it firsthand. Every department, every product, every roadmap meeting eventually funnels toward the same conclusion:“Can we add AI to this?”Hackathon after hackathon, teams are pulled together to “ideate” around AI, often regardless of whether the problem actually calls for it, or whether the people involved have any background or interest in AI at all.But beyond the buzzwords and demo days lies a far more uncomfortable question:What does all of this mean for software engineering jobs?What Generative AI Is and What It Isn’tGenerative AI, particularly large language models, are fundamentally probabilistic systems. When you type a prompt into a model like ChatGPT, each word in its response is simply the most statistically likely word to follow the previous one, based on massive amounts of training data.The result looks like reasoning. It sounds coherent.But there’s no actual understanding happening.Despite the popular narrative, we don’t even fully understand how human reasoning works, so the idea that scaling LLMs linearly leads to Artificial General Intelligence feels, at best, optimistic and, at worst, fantastical.What we have today isn’t a machine god.It’s more like an idiot savant, exceptionally good at mimicking logic without possessing it.And yet… that’s still enough to disrupt an industry.Why Software Is Especially VulnerableHistory gives us a useful analogy. In the 19th century, agriculture didn’t require humanoid robots to be automated. Tractors and harvesters, much simpler machines, were enough to replace massive amounts of human labor.Software engineering may be facing a similar moment.Most software written today is repetitive, predictable, and digital by nature—making it ideal training data for AI. Unlike other forms of engineering, software exists almost entirely as text, freely available in repositories across the internet.That makes it one of the easiest professions for AI to encroach upon.The Cracks Beneath the HypeThat said, the current generation of AI tools has real limitations:* Training data quality is inconsistent.Open-source code varies wildly in quality, age, and correctness.* AI increasingly trains on AI-generated code.This creates a feedback loop, a snake eating its own tail, where quality can degrade over time.* Context windows are limited.Large codebases don’t fit cleanly into an AI’s short-term memory, leading to hallucinations and subtle but dangerous bugs.* Critical work remains semi-analog.Requirements gathering, stakeholder negotiation, system design, and architectural judgment are not fully digitized, and therefore not easily learned by machines.The result? AI excels at greenfield code but struggles with large, messy, real-world systems.Fewer Engineers, Different RolesEven with these limitations, it’s hard to deny the trajectory. AI already produces code that is good enough most of the time and it will continue to improve.The likely outcome isn’t the total elimination of engineers, but rather:* Fewer roles overall* Greater emphasis on human judgment* More value placed on reasoning, intuition, and systems thinkingThis isn’t unprecedented. Aerospace and electrical engineering went through similar transitions long ago.Four Ways to Survive the AI TigerRight now, it feels like AI is a tiger chasing the tech industry. I see four possible coping strategies:1. Ride the TigerWork on frontier AI itself.* Pros: High pay, world-changing potential* Cons: Brutal competition, high barriers to entry, and the risk of an AI winter2. Outrun the TigerConstantly chase the newest languages and frameworks.* Pros: Short-term job security and high compensation* Cons: Career roulette, cognitive fatigue, and diminishing returns with age3. Tame the TigerBecome a generalist who orchestrates AI tools.* Pros: Massive leverage, small teams, long-term relevance* Cons: Everyone is trying this, and true generalists are rare4. Hide from the TigerWork in regulated or hard-to-digitize industries.* Pros: Stability for now* Cons: Digitization eventually comes for everyoneNo Perfect Ending, Only TradeoffsI don’t see a future where everyone in tech wins. But I do see paths where some engineers adapt, specialize, and survive.This isn’t optimism. It’s systems thinking.And yes, I know this sounds a little bleak but it’s my honest assessment after spending a long time thinking about the problem.That said, I’m just a laid-off ex–Big Tech flunky.You should take everything I say with a generous grain of salt.If you’re curious to follow along as I navigate life after Big Tech, thinking, building, and occasionally ranting, then feel free to subscribe and join the journey.Thanks for reading.Talk soon. Get full access to AsianDadEnergy's Newsletter at asiandadenergy.substack.com/subscribe