Comment by pico303
3 days ago
At least one CEO seems to get it. Anyone touting this idea of skipping junior talent in favor of AI is dooming their company in the long run. When your senior talent leaves to start their own companies, where will that leave you?
I’m not even sure AI is good for any engineer, let alone junior engineers. Software engineering at any level is a journey of discovery and learning. Any time I use it I can hear my algebra teacher telling me not to use a calculator or I won’t learn anything.
But overall I’m starting to feel like AI is simply the natural culmination of US economic policy for the last 45 years: short term gains for the top 1% at the expense of a healthy business and the economy in the long term for the rest of us. Jack Welch would be so proud.
> When your senior talent leaves to start their own companies, where will that leave you?
The CEO didn't express any concerns about "talent leaving". He is saying "keep the juniors" but he's implying "fire the seniors". This is in line with long standing industry trends and it's confirmed by the flowing quote from the OP:
>> [the junior replacement] notion led to the “dumbest thing I've ever heard” quote, followed by a justification that junior staff are “probably the least expensive employees you have” and also the most engaged with AI tools.
He is pushing for more of the same, viewing competence and skill as threats and liability to be "fixed". He's warning the industry to stay the course and keep the dumbing-down game moving as fast as possible.
Well that's even stupider. What do you do when your juniors get better at using your tools?
The 2010's tech boom happened because big tech knew a good engineer is worth their weight in gold, and not paying them well meant they'd be headhunted after as little as a year of work. What's gonna happen when this repeats) if we're assuming AI makes things much more efficient)?
----
And that's my kindest interpretation. One that assumes that a junior and senior using a prompt will have a very close gap to begin with. Even seniors seem to struggle right now with current models working at scale on Legacy code.
Maybe the head of AWS knows the current AI hype cycle is just that.
100%, and this is him selling the new batch of AWS agent tools. If your product requirements + “Well Architected” NFRs are expressed as input, AWS wants to run it and extract your cost of senior engineers as value for him.
Also fits in very well to amazon's famously low average tenure/ hiring practices.
I think AI has overall helped me learn.
There are lots of personal projects that I have wanted to build for years but have pushed off because the “getting started cost” is too high, I get frustrated and annoyed and don’t get far before giving up. Being able to get the tedious crap out of the way lowers the barrier to entry and I can actually do the real project, and get it past some finish line.
Am I learning as much as I would had I powered through it without AI assistance? Probably not, but I am definitely learning more than I would if I had simply not finished (or even started) the project at all.
What was your previous approach? From what I've seen, a lot of people are very reluctant about picking a book or read through a documentation before they try stuff. And then they got exposed to "cryptic" error message and then throw the towel.
I always used to try doing that. Really putting in the work, thoroughly reading the docs, books, study enough to have all the background information and context. It works but takes a lot of time and focus.
However, for side projects, there may be many situations where the documentation is actually not that great. Especially when it comes to interacting with and contributing to open source projects. Most of the time my best bet would be to directly go read a lot of source code. It could take weeks before I could understand the system I'm interacting with well enough to create the optimal solution to whatever problem I'd be working on.
With AI now, I usually pack an entire code base into a text file, feed it into the AI and generate the first small prototypes by guiding it. And this really is just a proof of concept, a validation that my idea can be done reasonably well with what is given. After that I would read through the code line by line and learn what I need and then write my own proper version.
I will admit that with AI it still takes a long time, because often it takes 4 or 5 prototypes before it generates exactly what you had in mind without cheating, hard coding things or weird workarounds. If you think it doesn't, you probably have lower standards than me. And that is with continuous guidance and feedback. But it still shortens that "idea validation" phase from multiple weeks to just one for me.
So: is it immensely powerful and useful? Yes. Can it save you time? Sometimes. Is it a silver bullet that replaces a programmer completely? Definitely no.
I think an important takeaway here also is that I am talking strictly about side projects. It's great as the stakes are low. But I would caution to wait a little longer before putting it in production though.
The biggest blocker for me would be that I would go through a "Getting Started" guide and that would go well until it doesn't. Either there would be an edge case that the guide didn't take into account or the guide would be out of date. Sometimes I would get an arcane error message that was a little difficult to parse.
There were also cases where the interesting part of what I'm working on (e.g. something with distributed computing) required a fair amount of stuff that I don't find interesting to get started (e.g. spinning up a Kafka and Zookeeper cluster), where I might have to spend hours screwing around with config files and make a bunch of mistakes before I get something more or less working.
If I was sufficiently interested, I would power through the issue, by either more thoroughly reading through documentation or searching through StackOverflow or going onto a project IRC, so it's not like I would never finish a project, but having a lower barrier of entry by being able to directly paste an error message or generate a working starter config helps a lot with getting past the initial hump, especially to get to the parts that I find more interesting.
> At least one CEO seems to get it.
> (…)
> I’m not even sure AI is good for any engineer
In that case I’m not sure you really agree with this CEO, who is all-in on the idea of LLMs for coding, going so far as to proudly say 80% of engineers at AWS use it and that that number will only rise. Listen to the interview, you don’t even need ten minutes.
> I’m not even sure AI is good for any engineer, let alone junior engineers. Software engineering at any level is a journey of discovery and learning.
Yes, but when there are certain mundane things in that discovery that are hindering my ability to get work done, AI can be extremely useful. It can be incredibly helpful in giving high level overviews of code bases or directing me to parts of codebases where certain architecture lives. Additionally, it exposes me to patterns and ideas I hadn't originally thought of.
Now, if I just take whatever is spit out by AI as gospel, then I'd be inclined to agree with you in saying AI is bad, but if you use it correctly, like any other tool, it's fantastic.
> When your senior talent leaves to start their own companies, where will that leave you?
In the case of Amazon with a shit ton of money to throw at a team of employees to crush your little startup?
Imagine you have shit ton of money but only agents that generate 10% bad code? You crushing or beating anyone..
Cause Matt comes from a technical background. Most CEOs dont.
The whole premise is just silly of thinking we don't need juniors is just silly. If there's no juniors, eventually there will be no seniors. AI slop ain't gonna un-slop itself.
You also risk senior talent who stay but doesn't want to change or adopt, at least with any urgency. AI will accelerate that journey of discovery and learning, so juniors are going to learn super fast.
That’s still to be determined. Blindly accepting code suggestions thrown at you without understanding them is not the same thing as learning.
>will accelerate that journey of discovery and learning,
Okay, but what about work output? That's seems to be the only thing business cares about.
Also, maybe it's the HN bias but I don't see this notion where old engineers are rejecting this en masse. More younger people will embrace it. But most younger people haven't mucked in legacy code yet (the lifeblood of any businesses).