Software engineers should stop treating stability as the same thing as safety, Brian Jenney argues in IEEE Spectrum’s careers newsletter. His point is aimed at a field where short tenures and abrupt layoffs have made career planning less like climbing a ladder and more like maintaining a fault-tolerant system.
Jenney wrote that the average software engineer stays at a company for roughly two years, about half the tenure of workers in many other knowledge jobs. The recent wave of layoffs has made that instability harder to ignore, he said, but it did not create the underlying pattern.
His advice is not to chase chaos for its own sake. Jenney argues that engineers should separate career risk into two buckets: risk involving the job itself, and risk involving how they spend their learning time.
Job security can hide technical decay
Jenney framed the argument through his own career switch. He said he left a secure community college job in his 30s, with a union and pension, to become a junior developer at 31. People around him, including his mother, considered the move reckless, he wrote.
His first software job was at a grocery retailer. Jenney said he liked the company and the people, but later realized that engineers doing comparable work were earning twice as much. In the San Francisco Bay Area, he also concluded that his skills were not growing fast enough. He left for a small startup, where he said he learned more in nine months than in the previous two years and doubled his pay.
From that experience, Jenney argues that engineers should take deliberate risks to work with stronger colleagues and harder problems. The mechanism is mundane and useful: better engineers expose weaknesses, model better habits, and pull coworkers into projects that build marketable skills.
He does not present job-hopping as a magic salary machine. Jenney wrote that switching jobs for money can make sense early in a career, but the payoff shrinks after several moves and the stress can mount. The more durable return, in his view, comes from roles that make an engineer better.
The risk of staying put, he argues, is less visible. An engineer can become essential inside one company while specializing in aging tools that employers elsewhere no longer value. That person may look safe until a layoff or resignation forces a trip back to the open market.
Trend bets are time bets
Jenney’s second category is the risk of spending time on technologies before their value is obvious. He described mainstream skills such as cloud services, ReactJS, and AI as difficult for software engineers to ignore without damaging their prospects. A backend engineer who refuses to learn cloud architecture, he wrote, is courting obsolescence.
The less certain bets are smaller trends. Jenney said he began learning about retrieval-augmented generation, or RAG, about two and a half years ago, when few people in his circle were discussing vector databases. RAG is now closer to mainstream, giving him what he described as an early-adopter advantage.
Most such bets will not pay off, Jenney cautioned. He said he is now making a similar wager on voice AI, which may not become mainstream. His case for the bet is practical: if it does take off, he will not be starting from zero.
Jenney’s conclusion is that short-term career risk can produce longer-term stability when it builds skills the market wants. Comfort is useful, but in his telling, engineers should keep asking whether they are still learning before assuming the safe path is actually safe.
This story draws on original reporting from IEEE Spectrum.
