alt Hacker NewsSoftware engineers can no longer neglect their soft skills
Comments
I've heard this "soft skills are the only skills that matter" thing throughout my entire career but these days this is indeed greatly amplified.
Our CEO started sharing screenshots of his xitter/linkedin feeds and most of them contain wisdoms like "Opus 4.5 is better than 90% of talent". There is also longer form business fan fiction. It usually follows this template: there is a hero - a humble manager/sales person, and a villain - a cocky, nerdy software developer. The villain claims that some task is impossible or takes months to complete. Then the humble hero (equipped with Opus 4.5) completes the task in 2 hours. The villain is then humiliated/fired and everyone lives happily ever after.
These posts definitely contribute to the declining morale among employees. Nobody goes "above and beyond" anymore - we just strictly doing the tickets and nothing more.
> Today, I use Claude Code for almost all non-trivial programming tasks and have spent $500+ on it just last December.
Truthfully, I don't want to get advice from people who become addicted to AI, sorry. The money investment that person did, already leaves me with tons of questions.
Be careful, engineers, when interacting with soft skill experts not to join their reality distortion field where it’s all about coordination, alignment, bizness strategy, clever planning. Whereas the real stuffs are just implémentation details, quickly solved.
In the end, they might convince you that 2+2=5.
The sort of people who have been able to neglect their soft skills are probably still going to be able to do so in the future. But those were always the extreme intelligence outliers, generational talent type figures.
For most software engineers, neglect of soft skills have always been a career tarpit that leads nowhere you want to end up. Being able to navigate social settings and to communicate well is a force multiplier. For most people, it really doesn't matter how good you are if nobody understands what you are saying and you can't convince other people to buy into your ideas. You far more often see moderately successful charlatans that are all talk than successful people with awful communication skills. Of course if you're able to walk the talk, that's when you can really go places.
So I opened this article to find out at the very beginning, that author put's a lot of money to AI providers... So probably also used it to write this article. So, according to the rule "text which is not worth of spending time on writing is also no worth of reading" I closed it.
I've had a long career in software and my conclusions is that if soft skills are valued over hard skills, the organization is already captured by talentless engineers and leaders. There are holdouts in the world, where execution is king, find those places and run away from soft skill fortresses. This is more true than even with LLM-amplified productivity.
It won't help. LLMs are good at soft skills, too. There's a whole "AI girlfriend" industry, and it's quite successful.
So, there is professional skills and then there is "soft skills" which at many places is a euphemism for political theatre. Nearly everyone I've ever met has professional skills; few can stomach "soft skills"
I thought this article was going to be about something else ...
It is really about prompting and writing specs - the "soft" (but really "hard") skill of giving detailed specs to an LLM so it does what you want.
I think the more important, truly soft, skill in the age of AI is going to be communicating with humans and demonstrating your value in communicating both vertically up and down and horizontally within your team. LLMs are becoming quite capable at the "autistic" skill of coding, but they are still horrible communicators, and don't communicate at all unless spoken to. This is where humans are currently, and maybe for a long time, irreplaceable - using our soft skills to interact with other humans and as always translate and refine fuzzy business requirements into the unforgiving language of the machine, whether than is carefully engineered LLM contexts, or machine code.
As far as communication goes, I have to say that Gemini 3.0, great as it is, is starting to grate on me with it's sycophantic style and failure to just respond as requested rather than to blabber on about "next steps" that it is constantly trying to second guess from it's history. You can tell it to focus and just answer the question, but that only lasts for one or two conversational turns.
One of Gemini's most annoying traits is to cheerfully and authoritatively give design advice, then when questioned admit (or rather tell, as if it were it's own insight) that this advice is catastrophically bad and will lead to a bad outcome, and without pause then tell you what you really should do, as if this is going to be any better.
"You're absolutely right! You've just realized the inevitable hard truth that all designers come to! If you do [what I just told you to do], program performance will be terrible! Here is how you avoid that ... (gives more advice pulled out of ass, without any analysis of consequences)"
It's getting old.
DjangoCon has a policy to not use the term "soft skills" and instead classify them as "professional skills" - communication and collaboration and management and leadership.
I think they're 100% right about that. There's nothing easy or soft about what gets classified as "soft skills".
> we won't be able to AI our way into better communication skills
Why not?
I always find these articles funny. There's someone almost triumphantly declaring that AI is able to take over the hard skills tasks from oh so dreaded engineers, but the authors can somehow not imagine that their soft skills - which are often they only ones they have - could be done by AI as well.
I used to work with a brilliant asshole - he had this babyface but would constantly belittle his fellow engineers.
Everyone hated him but the CEO loved him since he thought he was their golden ticket to the promised land.
At some point he went with us to a shooting range for target practice, immediately after he developed a fascination for guns, he never threatened anyone, but one day he buys a Kevlar vest and tells all his fellow devs.
The following Monday was his last day..
I'm just going to sharpen up my hard skills so that I don't have to suck up on my soft skills. If that doesn't work out as I wish, well, since I already have a job, and I'm already 40+, and it looks like my company doesn't go away in a few years and we are so short-handed, I think I'm fine.
Just to clarify that I'm not a jackass in real-life. In fact, I'm perfectly OK with all sorts of soft skills -- after all, my current position requires me to do so. But I just try to maintain a minimum level of soft skills to navigate the shoreline -- not interested to move up anyway.
when could we?
The irony here is that universities are struggling to teach writing skills, due to massive cheating with AI.
This couldn't ring more true to me - I think one of the consequences of the rapid change in the profession we are seeing is that skills that typically were required only at more senior levels become required further down the stack.
If I was a junior today, I'd be studying business impact, effective communication, project management, skills that were previously something you could get away with under-indexing on until senior+.
Unless you're buried deep in an organization, the main value that programmers provide is to help translate the needs of a business into a program that, when compiled, creates a reliable artifact that, when executed, works reliably and is easy to use.
Writing code is just how that happens, sometimes. Soft skills are essential to communication with the users and product managers.
The classic "technically strong but lacking soft skills" unheard story about you through the office after your results.
What about those professional software developers still refusing to use AI / LLMs? I know a couple and they're still churning out code completely 100% manually.
Heck, I even know a guy who refuses to use an IDE with Java and the indenting is a mess, but he gets there.
The claim in the article that AI is good at writing code, is that claude code is written by claude code.
calculators doing a perfect job did not end accountants' jobs, it made it faster.
This has never been the case for those of us doing consulting, soft skills are a must have requirement when dealing with customers on regular basis.
I would submit that this could be based on a stereotype that “coder = antisocial.”
Why are we assuming that people who write code don’t have soft skills?
The youngest generation who joined the profession are probably in it for the the salary versus the older generations who came from computer clubs and dungeons and dragons groups of the 1970s/1980s along with a culture where having a niche interest was socially ostracizing and uncool.
I wonder if the youngest generation entering the profession is much more of a cross section of regular people?
Counter-point, a three person dev team displaces a multi million revenue company.
Why have a slow human CEO when machines are faster..
"I deal with the goddamn customers so the engineers don't have to. I have people skills! I am good at dealing with people, can't you understand that? WHAT THE HELL IS WRONG WITH YOU PEOPLE?!"
Thank Open Source and Free Software which commodified hard skills and pushed business to value them as zero.
You never really could. If you hear of a (very) succesful software engineer with horrid soft skills they’re a 1%er chance wise
I think AI coding agents will quickly pick all the low hanging fruits and plateau.
Everything can be vibed will be vibed until everyone hits a wall, where no docs to form corpus nor instructions for prompts exist. There are problems that are yet to be named, but how can you name things when humans aren't the one to experience patterns of a thought process?
And naming things is one of the only two hard things in computer science
In education, I view them as life skills, durable skills.
In a team environment, half of the job is communication.
That LLMs do a better job if you know what you are asking for is old news.
But to be honest, I usually don't care to write properly into an LLM prompt. An LLM will ignore grammar and form and just extract the essence. If I make an actual mistake I will notice quickly and fix it. If I'd send slack messages like that to an peer, they'd either mock me or simply think I am dumb. We also know the stories about people that use LLMs for any communication or anything they write. Probably for the exact reason that being lazy with writing is acceptable now. My call is that writing skill will decline, not improve. This could probably be the case for anything that people use LLMs as a proxy for.
Do we accuse AI of having poor soft skills when it doesn't do what we want?
I think the problem is knowing how to bridge knowledge gaps. That just comes from experience and there are no shortcuts on either end of the gap.
Empathy does matter a little bit, but to focus so much on it is plain neurotic. Consider how much less friction there is when the interactions can be kept brief. Everyone is already familiar with the various situations and problems that can arise (like on a sports team). That's pure hard skills, not soft skills.
Posts like this are flamebait for the extreme ends of these gaps: stubborn mediocre programmers and arrogant dumb management.
Watch them
My take:
Businesses have valued "cohesion" over "correctness" for some time (at least the last 10 years of my career) with the thinking that they can always eventually get to a correct solution, but teams that aren't cohesive do not work toward the goal they fight amongst themselves until they tear themselves apart (as a former Python dev I have seen teams that have one or two members fight for MONTHS over which set of linters to use)
I also want to say that the only source of "bugs" is misunderstandings - of what the technology does, what the business wants, or what the customer wants (two thirds of that is "soft skills"). We've created DDD to try and address one third of those potential issues, but we're not there yet.
Counter point - If one person can now do the work of an entire team the level of communication skills required will actually be simplified.
So now instead of needing to manage multiple stakeholders and expectations of 10 different middle managers you'll probably just have a 1:1 with a single person.
Poppycock. If a software engineer truly solved a hard problem they they would be a billionaire overnight.
Now that everyone has a assistant that can work 24/7, talk to customers, get requirements There is no excuse for neglecting hard skills.
What this article calls "soft skills" is largely just experience which is often wrong in tech.
As usual - the advice is essenially rats from a sinking ship all the way. “You all need to do this narrow thing to survive now”.
2016 to truckers: “Learn to code LOL”
2026 to coders: “Learn soft skills”
https://news.ycombinator.com/item?id=46436872
Look, I personally am taking full advantage of exactly the skills described. I was the one who posted the above thing on HN showing how I am 20-50x more productive now, complete with a 4 hour speedrun video. I usually try not to just talk and point out current problems, but build solutions AND show (github, youtube) with specific details so you can watch it and apply it for yourself. But I am telling you:
1) most people will not adapt, so we will need UBI for those who don’t
2) eventually even those who adapt will be replaced too, so we will need UBI for everybody
It is after all a thin layer that remains. I remember Kasparov proudly talked about how “centaurs” (human + machine working together) in chess were better than machines alone… until they weren’t, and human in the loop became a liability.
But the problem is more widespread in the last 70 years. Just look around. Industry always tells the individual they can do some individual action downstream to clean up the mess they create upstream, and it is leading the entire planet into ruin:
https://magarshak.com/blog/government-and-industry-distract-...
In fact, the human population in modern environments has been living large on an ecological credit card and the bill is coming due for our children, because all the “individual responsibility” stuff — where you can somehow diet, exercise and recycle your way out of things corporations do upstream — is all a gient lie and always has been. So the negative externalities just build up until the next generation won’t be able to ignore them anymore, but it could be too late. Whether that’als day zero for water in cities, or factory farms for meat with antiobiotic resistance, or fossil fuels and greenhouse gases to subsidize the car industry, or ubiqitous microplastic plastic pollution around thr world (yes, personal plastic recycling was just another such scam designed to keep you docile and not organize to force corporations to switch to biodegradeable materials.) The “anthoposcene” is seeing a decline in insects and all species of animal except humans and farm animals. Coral reefs are bleached, kelp forests and rainforests are decimated, and governments work with industry to eg allow Patagonian forests to be burned for new developments and then smokey the bear says “only YOU can prevent forest fires”. Think about it.
Software 'engineers' are going to have learn the hard skills that they have gotten by without thus far. The ones who were never engineers in the first place will have to learn 'soft skills' because they are incapable. I believe @ Tsoding put it best: The only thing AI will do, in so far as coding goes, is to remove a lot of people who should never have been in it in the first place.
[dead]
[dead]
[dead]
I don't think it has ever been the case that you could neglect soft skills. You will hear this over and over, in every area of every business: people become successful by adjusting their behaviour to what works for the business. Sometimes this is called being a slick politician, sometimes it is called avoiding getting bogged down in politics.
But it's never been the case that a dev could just focus on technical things and not spend any time figuring out the context they are working in, and behaving accordingly.
My first day of work, this is what my boss said to me: "Look at this trading floor. There's screens everywhere, everything is numbers. Deltas, gammas, vegas. Everything is calculated by computers. But don't forget, every business is a people business!"