It also seems possible that clever, young outsiders will, by creatively employing AI, displace senior devs. Because unlike with law and accounting, novices in software development can generate a lot of their own reps at low cost (mostly just their time, which will be cheap compared to senior devs).
Another possibility could be using junior devs (with AI) to come up with enormous numbers of fresh ideas and minimally working programs, with the promising ones then worked on with senior devs.
Most fundamentally, because AI lowers the cost of testing out ideas and potential software products, it seems like a tool to test notions of what you should build or continue to offer customers in the first place.
The real threat is SPAM. No matter how great you are at spotting diamonds in the rough, if the rough expands exponentially (unlike the singularity), the real trouble is going to be about how to actually evaluate whether or not whatever is presented is a solid code, or three AIs in a trench coat.
We already see it overwhelming HR systems, but once all this code gets generated, how is anyone going to know, without being extremely competent, that these vibe coded things are sufficient demonstration of future value.
Or worse, how are you going to even find these new age AI wunderkind, when it used to be hard to polish up a project and now anyone can spit shine their project.
I'm currently in the hiring process.. and my head is spinning..
1. Interviewing is becoming more difficult. Many skills we valued two years ago are genuinely becoming less valuable. 2. The skills we tested 2 years ago were in-part a proxy for evaluating critical think and systems thinking skills. So we need to re-evaluate our technical interview process. 3. It's genuinely less friction for me to prompt Claude than some of my SEII colleagues. And wtf - Claude is getting so good that it's starting to feel like it's outpacing the intellectual competency of some people. Sure - it does weird things like add Sleep instead of proper concurrency. SEIIs did that too, and we couldn't as easily reprogram them with Skill.md 4. Core competencies remain necessary. Systems thinking. SOLID principles. Communication skills. These skills are more important than ever. 5. Companies that offshored engineers and traded core competency for perceived-throughput are doing the same calculus with Claude. 6. Core business models are threatened. There is fear that revenue streams will dry up. How does one hedge against that risk? Humans are expensive. 7. Navigating this situation is hard and uncertain
I guess this is a problem if a company just wants to hire “some coders.” “Coders” is a term I’ve only heard used by those that don’t understand what’s involved in building and maintaining nontrivial software systems. You used to be able to get a high paying job just throwing together basic web pages in 1999. Non-technical folks have had site design tools for awhile that replace a lot of this work. There are more true software engineering jobs than there used to be and I think this trend will continue as LLMs create even more demand for software.
This is a real issue in legal as well. Now that large language models are creeping into big law and in-house environments, there's a concern that junior associates aren't getting the opportunity to take a step back and learn to consider the broader context.
When your bidding can be drafted and a path laid in microseconds, it's super easy to start off in the absolutely wrong direction. But you don't know you're headed the wrong way unless you've learned it. Unlike software where bugs like this are sometimes surfaced immediately via the compiler or interacting with the product, legal bugs are latent and only reveal themselves after the potentially massive damage is done[1].
These things are a massive unlock for well-trained senior lawyers who can spot the issue upfront. On the other hand, they amplify juniors' ability to introduce errors at the same time they deprive them of necessary understanding. Having a judge rule on a bad contract idea "at runtime" is a catastrophic failure mode.
[1] As an example of this, consider how Gary Kildall arguably flubbed the deal of a century when he allowed the DRI team to attempt to negotiate the IBM form non-disclosure agreement: https://tritium.legal/blog/redline
Halfway through this blogpost, I realized it's written for idiots used to the Buzzfeed/Twitter-style of "One sentence per paragraph". Fucking infuriating.
Not at all! That is my drafting method creeping through. I try to write each sentence as a separate paragraph first to make sure I don’t get too wed to a paragraph and can delete/move liberally. Then I try to go and merge the paragraphs that have cohesion. Seems like I didn’t do a great job here. Sorry it distracted you, but honestly thanks for giving it a chance and the feedback. Will do a better job on the next one.
Saw this play out in audit firms. They have a very rigid yearly cycle with each group coaching the next. It’s a good model because it’s basically hands on 1:1 tutoring on real world work.
Then leadership couldn’t resist the $$$ temptation of outsourcing the bottom tier to India.
Suddenly you’ve got people that have the senior title but can’t lead because they don’t understand the task they’re supposed to provide leadership on
And because it’s such a relentless yearly cycle you can’t do much to fix it. A single year of that caused substantial lasting damage
Companies like this suck to work at and will always exist. I started my career at one and was depressed because I thought it represented the broader software engineering job space. Find a better company is my advice.
I saw my doctor using chatgpt. But he's experienced, finished his schooling before all this. I wonder about the next generation of people who are supposed to save lives and build bridges
It's ironic that Microsoft execs don't worry about AI replacingtheir jobs. If that doesn't paint a very clear picture of why most execs keep repeating this trope, I'm not sure what is.
> ... arguing that senior software engineers must mentor junior developers to prevent AI coding agents from hollowing out the profession's future skills base.
Nah, I want that COBOL expert treatment in my senior years. Screw the profession, it ain't communism.
It also seems possible that clever, young outsiders will, by creatively employing AI, displace senior devs. Because unlike with law and accounting, novices in software development can generate a lot of their own reps at low cost (mostly just their time, which will be cheap compared to senior devs).
Another possibility could be using junior devs (with AI) to come up with enormous numbers of fresh ideas and minimally working programs, with the promising ones then worked on with senior devs.
Most fundamentally, because AI lowers the cost of testing out ideas and potential software products, it seems like a tool to test notions of what you should build or continue to offer customers in the first place.
The real threat is SPAM. No matter how great you are at spotting diamonds in the rough, if the rough expands exponentially (unlike the singularity), the real trouble is going to be about how to actually evaluate whether or not whatever is presented is a solid code, or three AIs in a trench coat.
We already see it overwhelming HR systems, but once all this code gets generated, how is anyone going to know, without being extremely competent, that these vibe coded things are sufficient demonstration of future value.
Or worse, how are you going to even find these new age AI wunderkind, when it used to be hard to polish up a project and now anyone can spit shine their project.
I'm currently in the hiring process.. and my head is spinning.. 1. Interviewing is becoming more difficult. Many skills we valued two years ago are genuinely becoming less valuable. 2. The skills we tested 2 years ago were in-part a proxy for evaluating critical think and systems thinking skills. So we need to re-evaluate our technical interview process. 3. It's genuinely less friction for me to prompt Claude than some of my SEII colleagues. And wtf - Claude is getting so good that it's starting to feel like it's outpacing the intellectual competency of some people. Sure - it does weird things like add Sleep instead of proper concurrency. SEIIs did that too, and we couldn't as easily reprogram them with Skill.md 4. Core competencies remain necessary. Systems thinking. SOLID principles. Communication skills. These skills are more important than ever. 5. Companies that offshored engineers and traded core competency for perceived-throughput are doing the same calculus with Claude. 6. Core business models are threatened. There is fear that revenue streams will dry up. How does one hedge against that risk? Humans are expensive. 7. Navigating this situation is hard and uncertain
I guess this is a problem if a company just wants to hire “some coders.” “Coders” is a term I’ve only heard used by those that don’t understand what’s involved in building and maintaining nontrivial software systems. You used to be able to get a high paying job just throwing together basic web pages in 1999. Non-technical folks have had site design tools for awhile that replace a lot of this work. There are more true software engineering jobs than there used to be and I think this trend will continue as LLMs create even more demand for software.
This is a real issue in legal as well. Now that large language models are creeping into big law and in-house environments, there's a concern that junior associates aren't getting the opportunity to take a step back and learn to consider the broader context.
When your bidding can be drafted and a path laid in microseconds, it's super easy to start off in the absolutely wrong direction. But you don't know you're headed the wrong way unless you've learned it. Unlike software where bugs like this are sometimes surfaced immediately via the compiler or interacting with the product, legal bugs are latent and only reveal themselves after the potentially massive damage is done[1].
These things are a massive unlock for well-trained senior lawyers who can spot the issue upfront. On the other hand, they amplify juniors' ability to introduce errors at the same time they deprive them of necessary understanding. Having a judge rule on a bad contract idea "at runtime" is a catastrophic failure mode.
[1] As an example of this, consider how Gary Kildall arguably flubbed the deal of a century when he allowed the DRI team to attempt to negotiate the IBM form non-disclosure agreement: https://tritium.legal/blog/redline
> https://tritium.legal/blog/redline
Halfway through this blogpost, I realized it's written for idiots used to the Buzzfeed/Twitter-style of "One sentence per paragraph". Fucking infuriating.
Not at all! That is my drafting method creeping through. I try to write each sentence as a separate paragraph first to make sure I don’t get too wed to a paragraph and can delete/move liberally. Then I try to go and merge the paragraphs that have cohesion. Seems like I didn’t do a great job here. Sorry it distracted you, but honestly thanks for giving it a chance and the feedback. Will do a better job on the next one.
I liked it, one idea to a sentence is one of the best pieces of writing advice one can follow. The article reads like a script, it's not bad at all.
Instead of one paragraph per sentence, try one sentence per line.
https://sive.rs/1s
Saw this play out in audit firms. They have a very rigid yearly cycle with each group coaching the next. It’s a good model because it’s basically hands on 1:1 tutoring on real world work.
Then leadership couldn’t resist the $$$ temptation of outsourcing the bottom tier to India.
Suddenly you’ve got people that have the senior title but can’t lead because they don’t understand the task they’re supposed to provide leadership on
And because it’s such a relentless yearly cycle you can’t do much to fix it. A single year of that caused substantial lasting damage
Companies like this suck to work at and will always exist. I started my career at one and was depressed because I thought it represented the broader software engineering job space. Find a better company is my advice.
I saw my doctor using chatgpt. But he's experienced, finished his schooling before all this. I wonder about the next generation of people who are supposed to save lives and build bridges
Last time I checked they were HOPING for that. The times change I see.
It's ironic that Microsoft execs don't worry about AI replacingtheir jobs. If that doesn't paint a very clear picture of why most execs keep repeating this trope, I'm not sure what is.
> ... arguing that senior software engineers must mentor junior developers to prevent AI coding agents from hollowing out the profession's future skills base.
Nah, I want that COBOL expert treatment in my senior years. Screw the profession, it ain't communism.
I guess you didn't see https://claude.com/blog/how-ai-helps-break-cost-barrier-cobo...
See what? Where are the clients' success stories?