Coders: Will A.I. Take your Job?
Software Engineers are getting laid off and have questions for ChatGPT
Hey Everyone,
On this publication, there will be times I write several times a week, and other times I will go 10 or even 14 days without a Newsletter, that’s just the way things are. Lately in the News I’ve been reading a lot about the layoffs in Technology how many software engineers are being impacted.
With GitHub Copilot and Google’s PaLM tool and many other auto-complete like A.I. assistants, they say the jobs and roles of coders won’t be disrupted, but can we trust ChatGPT for the future of coding? I’d like to seriously ponder the issue.
Matt Welsh, a former professor of Computer Science and Google engineer believes that programming jobs, as we know them today, will cease to exist in three years. If ChatGPT and GitHub Copilot improve at the rate they have been improving, won’t GPT Agents realize a lot of coding tasks in the decades ahead?
Thenewstack writes:
Welsh thinks programming is on the cusp of changing from a job that humans do, to one that robots will do, thanks to technologies like ChatGPT and Copilot. In his view, programmers will need to evolve into “teachers” of AI programs — or perhaps product managers or code reviewers, the two human roles he thinks are relatively safe from the robots. But don’t expect to continue your career as an actual programmer, he says, because machines are taking over that role.
Listen to Matt Welsh here.
Just as software development has slowly become more inclusive between the genders, is there actually the possibility it will begin to disrupt these highly skilled knowledge workers?
Welsh, who has held senior engineering positions at Google and Apple, had written an article on this topic for the January 2023 edition of the Communications of the ACM magazine.
I’m fairly interested in how A.I. impacts the knowledge workforce and the future of engineering. If you consider that expensive engineers and their salaries is a huge bottleneck for a lot of startups and innovation! This has incredible implications across industries, if it’s true.
Some are so blinded by ChatGPT hype they say a lot of things, and make some outlandish convictions, predictions and have a weird solidarity with the Venture Capital and BigTech folk who benefit the most.
But we also have to consider the factual possibility?
If A.I. learns to code itself, what can it become?
What could be the impact on coding jobs and demand for actual software engineers, data scientists, data engineerings, MLops, analytics folk, business intelligence and IT Workers broadly speaking and so forth?
While we can ruminate on the top programming languages for A.I. development, will our jobs exist in 5 or 10 years? We may have to pivot in ways that will change over time.
People like Matt Welsh believe that the only thing limiting co-pilot today is more data and compute. Is that true?
Software Engineering Jobs Are not Safe in 2023
Insider wrote a hit piece recently where they said:
Software engineers have overwhelmingly faced layoffs this year, per data from Revelio Labs.
Coders made up 14% of employees at tech companies, but represented 20% of layoffs in 2023.
It's a big shift from data from late last year, which showed recruiters were overwhelmingly cut.
That to me seems like a high number. Even a flashy new job like a prompt-engineer might not exist in three years time as GPT Agents evolve. A lot of the early builders in the Generative A.I. space will be made obsolete by new GPT and foundational models. They are gambling with innovation, at a time when the “speed of A.I.” is fast and peculiar.
Data from the Fall of 2022 pointed to software skilled folk being expendable in the “year of efficiency” in Silicon Valley.
According to data from Revelio labs.
Share with a friend:
How could some of our most in-demand knowledge workers actually be disrupted by these tools? Maybe GitHub Copilot just gets better and better much faster than even Microsoft expected? How could this be?
Friend or foe: Can computer coders trust ChatGPT?
Is this the End of programming? I mean, it makes for a good headline but really? Revelio seems to have gotten their data from layoffs.fyi (a site I mostly check for company wide layoffs).
"Earlier layoffs were focused on future hiring. Recruiters and HR were overrepresented in those earlier layoffs," said Ayas, who led the recent study. For these findings, Revelio pulled data from layoffs.fyi and Parachute List last Thursday.
With margin compressions occuring in Earnings and a slowdown in things like Cloud, PC Sales and Advertising, I guess it’s only inevitable for companies who over-hired in the near zero interest decade of the past.
Think of it this way, while layoffs were previously focused on the future of hiring, by targeting engineers, companies are now focusing layoffs on current business priorities and product priorities, which makes sense. But it’s also a bit scary. Retail and media layoffs I understand, but software engineers? It could be the dawn of a new era.
Were were always told engineering had a bright future full of demand. But at what point is this no longer true, is it some hypothetical 2028? Depending on the speed of A.I.?
The choice of programming language can affect an AI system's performance, efficiency, and accuracy. But what if A.I. itself masters programming relatively easily, and relatively quickly, will it still just be a “co-pilot”? What has Microsoft done?
A lot of coders use a combination of ChatGPT and GitHub Copilot and other tools. It does not mean they will be replaced, right?
Software Engineers are amongst the most common roles vulnerable to the tech layoffs of late 2022 and 2023, and likely some of 2024.
If Recruiters and Customer support are the most impacted, Software engineers aren’t exactly so far behind:
If Tech needs to save cash due to deteriorating margins, it really does make sense:
Maybe it’s not news to some of you, but I found it a bit surprising.
Are Coders the Odd Ones out due to A.I.?
There is one quote that really struck me: (the idea that this is a new normal).
“Before it was just, ‘the more or better, whatever it takes to recruit the best talent,’” said Daniel Keum, an associate professor of management at Columbia University’s business school, said of tech companies. “Now they’re cost-conscious, they want to be optimized and economized.”
He added, “That’s an entirely different mindset. We haven’t seen this before.”
The Tech Bleed Will Continue in 2023
The latest monthly jobs report by tech industry association CompTIA found that even though employment at tech companies (which includes all roles at those companies) declined slightly in March, employment in technical occupations across industry sectors increased by nearly 200,000 positions.
The average software engineer base pay in the US is $90,000, according to PayScale, but can be substantially higher at tech firms like Facebook, where such workers also get bonuses and stock options. The get rich move to the Valley card may be coming to an end.
Blame ChatGPT and declining margins. Seriously.
This research builds on data Revelio published with Insider's Aki Ito last September, which found that recruiters were overrepresented in layoffs by nearly 8%, compared to other positions. Then, software engineers were overrepresented in layoffs by nearly 4%.
If ChatGPT and Bard and GitHub Copilot are only getting started with Code! It could be a weird world for many knowledge workers in the Tech umbrella. Sorry Tech bros, your time is up! I’m kidding but I found all this data quite interesting and maybe salient to your world and your job?
I’ll stop because I’m rambling but this topic isn’t over, this is just the beginning.
I have mixed feelings about whether programmers will be completely replaced by AI. It is certainly on the cards that given a large enough model, a lot of programming jobs could be automated. As you say in the post, humans might just be reviewing the code at that point, so knowledge of coding would still be valuable but more from the point of view of reading. However, I also wonder if this will lead to a point where humans are not experienced enough to be able to catch bugs in such generated code? I mean, a senior programmer is so good at their job because they have spent several years writing code, they have seen and fixed a ton of issues in production so they know what things to look for when reviewing code. This would be a lost art.
For really massive, complex and critical projects, such as the Linux kernel or a database engine, I think we will still need skilled programmers to continue their work. Such projects are extremely complex and large for an AI model to meaningfully take over.
The other question is of innovation. Sometimes when faced with an engineering problem, humans come up novel ways to solve it in order to meet the constraints of the domain, this might mean inventing a new data structure, a new algorithm, or using an existing technique in novel way. I'm not very sure if these models can come up with such ideas, they would tend to stick to what they have seen in their training data.
I think re: the representation of different types of employees in layoffs it's important to normalize by the percent of the workforce made up by those employees. So instead of the percent of layoffs made up by software engineers, it would show the percent of software engineers laid off -- because tech companies are SWE heavy, so it makes sense that a uniform layoff across employee types would still be made up mostly by SWEs.