What are some specialties to look into for an older programmer?

December 28, 2021 9:24 AM

I'm looking to invest over the next year in another specialty, either on my own, or with a new company. Sort of a self-made masters degree. What is likely to be successful? Go wild, the stranger the idea the better!

Every few years, I enjoy switching fields within programming. So far, I've done early data engineering, manufacturing and hardware integration and QA, prototyping, software QA and automation, and now front end frameworks, UX, and a little bit of full stack. This has been across a half-dozen product focuses, like datacom and medical.

Things I'm looking for, understanding no role will tick all the boxes:

* Being a primary contributor. Or, phrased another way, likely to be respected and painful to lay off.
* Lasts ~10 years as a field, but not necessarily in one job.
* A reasonable number of jobs in the field
* Leverages my existing skills to some extent. I still like programming.
* Not too terribly soul-crushing. I've seen jobs that were automating the rejection of insurance claims... I couldn't handle it.
* Preferably, target companies that understand making software. A technical job track would be a dream.

Things I'm not looking for:

* Being self-employed. Been there, and it's not for me.
* Consulting. Tried it, the constantly selling gets me down.

Things I've considered:

* Data science - seems a little low on the primary contributor scale, but I like the work
* Machine Learning - I have concerns about the longevity of the field, and whether it will simply be taken over by tooling.
* Management/leadership - not my usual gig, but I've learned a lot over the years, it might be fun to share. Getting in the door as a manager with no prior experience seems problematic.
* Back-end/cloud specialization - probably the easiest to leverage with my existing skill set, very generalizable across companies, but, eh, not exciting. I like users.
* Double-down on front end - lots still to learn, and I could job hop pretty easily.

What else is there? Thanks all!

posted by SunSnork to Work & Money (6 answers total)

My current perspective on this comes from working at a large tech co., though I’ve worked at smaller orgs too.

Re: data science, our hiring process is pretty serious about the “science” part and expects incoming DS candidates to have PhDs. It’s not just stats-on-a-mac and it’s definitely not treated as a branch of engineering. On management and leadership, larger orgs with well-defined career ladders will let you move back and forth. One of my interviewers was a senior engineer who had switched from director-level leadership. I have a number of current coworker managers who’ve switched from engineering roles. The expectations on each side are different but there’s a defined pay scale equivalence. Also when you get to super-senior engineering levels your success is measured in terms of influence over larger groups. “You can’t type fast enough to succeed” by coding at the senior engineering levels is how a past boss of mine put it.

The two hardest disciplines out there are front-end and infrastructure, I think. You seem more interested in having users so I would go hard on the former. Its difficulty comes from staying on top of whatever browser makers are doing, constantly-shifting framework fads, and strictly-increasing user expectations for speedy and responsive user interfaces. The job market will probably look like a mass of early- and mid-stage startups trying to build or improve upon minimum-viable products while managing looming technical debt from past hasty engineering decisions.
posted by migurski at 10:36 AM on December 28, 2021 [3 favorites]

I'm an older programmer and I specialize in embedded systems with an IoT focus.

There's tons of work available. There are a myriad of companies that want to upgrade their technology and make it cloud-connected. Fields you wouldn't even think of day-to-day: kitchen equipment, power tools, test equipment, etc. Is it all useful? Hell, no! Do the companies want to do it and will pay for it? Hell, yes. Let the marketing goons figure out what is useful and sellable.

Having full-stack experience is also a huge plus here. If you can take an embedded Linux system and drive all the bootloader, kernel, application, and cloud/database/backend work by yourself you can succeed nearly anywhere. You might have to deal with some pretty ugly and old code at times, but that can be a fun challenge too.
posted by JoeZydeco at 10:41 AM on December 28, 2021 [1 favorite]

As another poster commented, there are a lot of opportunities now, and there seems to be a shortage of qualified people. I'll probably take some flack for saying this, but I think it's worth saying anyway: there are so many problems facing humanity that I'd like to suggest looking for ways of putting your good skills and experience into building beneficial systems instead of (e.g.) useless gadgets that some marketing people think will make a buck. Here are some ideas off the top of my head, but there are undoubtedly more if you start thinking along these lines:

• Development of renewable energy production like wind and solar are predicted to soar over the next 5 years (article in MIT review, article in NYT). Making these systems work, and work efficiently, needs computerized systems. Related areas include energy storage, energy transportation, forecasting energy needs, etc. Needless to say, helping this sector would help the fight against climate change.
• Much of the US has poor internet access because the infrastructure to reach rural areas is missing, and building it is challenging due to a number of factors. Getting it built will require creativity. Companies like Microsoft and many small telecoms are trying to change that (e.g. via Microsoft Airband). It would make a huge difference to many people's lives to have better (in some cases any) internet access. There's probably a wide range of opportunities here, in many regions of the country.
• You mentioned machine learning – there is a lot of interest in applying ML to helping face climate change. I mentioned renewable energy above; that's an area where machine learning is being applied, but there are others. A recent academic paper by some leading researchers provides a survey and may help spark some ideas; a McKinsey report on applying AI for social good has even more (see their figure 1 for a summary).
• We are past the point of being able to stop climate change; the only question is how bad it will be. Another line of thinking is, what kinds of systems do we need to build now to help future people during emergencies and disasters? Among them are things like robust communications systems for emergencies (low-power, distributed, cheap, etc.), off-the-grid self-contained data access points with medical databases and instructions for emergency care, etc.

Again, this is just meant as a starting list, something to spark some ideas. All of these areas will be active for at least 10 years, and there are both mature companies and startups involved.

(Some people may rightly point out that technology alone will not help us, but this was a question about working with technology, so …)
posted by StrawberryPie at 6:44 PM on December 28, 2021 [3 favorites]

Umnn, perhaps you've heard of the metaverse? The B2B market is probably bigger than consumer games and apps. Just one example is Carbon Origins which uses people in a VR headset to supervise delivery and agricultural robots.
posted by Sophont at 2:45 AM on December 29, 2021

My last job before retirement was with a small company in the insurance industry with about 50 employees. They called me the CIO when it suited their fancy, but in fact, i was the only permanent IT employee. I didnt have any management experience, but obviously that was not a big issue. A consultant managed the network, PC issues, etc. t was 2005, and a client/server situation, so I only dealt with the internet for sending and receiving files.

It was great. I did the programming, answered questions for audits, and a miscellany of other tasks. By the time I retired, there were two or three other IT employees.

My point is that if the company is small enough, management isn't just management. You get some control over which stack, which cloud, what quality control and testing, etc. Life can be good.
posted by SemiSalt at 6:01 AM on December 29, 2021 [1 favorite]

Purely FYI, as people like me retire or die off, there will be an increasing need for people who know COBOL both for maintaining systems and converting off of them to more modern languages. FWIW.
posted by forthright at 9:12 AM on December 29, 2021 [2 favorites]