Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

LLMs are eroding my software engineering career and I don't know what to do

human-in-the-loop.bearblog.dev

Post reply on HN
Print view View original

871–880 of 1001 posts

Re: LLMs are eroding my software engineering career and I don't know what to do

#871

Post by dinkumthinkum »

goosejuice wrote: post #215
dgan wrote: post #64

Earlier quoted context omitted.

Lol thats brave on your part, given that a mistake can cost thousands and you have no accountability (punch!) from an LLM

In the US, all you need to work in tax prep is a high school diploma and most individuals are not worth the cost of an audit. I wouldn't say it's particularly brave, in fact LLMs are probably better at identifying mistakes than most tax payers. The % of Americans using a CPA to file taxes is fairly small.

If you go by percentage of Americans, well consider the percentage of Americans that are actually net taxpayers.

Re: LLMs are eroding my software engineering career and I don't know what to do

#872

Post by theshrike79 »

chasd00 wrote: post #768
theshrike79 wrote: post #665

Earlier quoted context omitted.

This is the way to do it. You can have the Big Boy LLM do _everything_. It can and it will do it. It will also cost fucktons of money and take a long time. But if you build tools (with AI) that do as many tasks in the process deterministically as possible and let the AI use those, it'll be a lot faster and cheaper to run it. As a bonus you can eventually drop the expensive cloud AI and run a small/medium sized local…

You know, i think harnesses and tools are the next "webapps" for the industry. Everyone is going to be making their own. Some will be great some will be meh but i think that's where things are headed.

Yep, there are actual papers written on how you can have a mid-tier model punch way above its weight with an optimised harness and toolset.

Claude is still the best commercially available harness IMO, pi.dev is super good but not something I'd give to non-enthusiasts or would recommend using in an enterprise environment.

I see companies writing their own custom harnesses on top of opencode/pi.dev/crush later down the line. Instead of having a set of skills or MCPs you can just have all of the default stuff built in and automatically updated via normal IT workflows.

Re: LLMs are eroding my software engineering career and I don't know what to do

#873

Post by marekful »

I think LLMs are the great filter of software engineers. They can do a lot reasonable well, but they lack the ability to be an overarching architect of complex projects (and it's my conviction that they can never achieve that).

Only the best humans with insight, intuition and pattern recognition and application in non trivial scenarios can fill that gap.

Re: LLMs are eroding my software engineering career and I don't know what to do

#874

Post by sensanaty »

jalev wrote: post #41
iandanforth wrote: post #18

Wut? I pilot LLMs all day but there's no way in hell I'd agree to be at the helm of a finance product. That first pillar is still there. Maybe the author isn't aware of the impact they have, but I know, with the evidence of reverted PRs, that when I step outside my area of deep knowledge I can no longer call BS on the agents. Our most capable agent, with access to the same kind of distributed systems the author talks…

Unfortunately every software related industry is embracing LLM/Codegen. Your banks, fintechs, insurance. Everyone. Your concerns are the same I'm having, yet it's regularly dismissed or hand-waved away as "don't worry about it the delivery velocity/ROI is worth it"

I work in a large, established fintech and we're in no rush to shove LLMs everywhere. We have access to all the AI tooling we'd ever need with basically no limits, but AI or not, you're responsible for what you ship, so most people are less YOLO about the whole thing. From the quarterly surveys, most people in the co view it as a slight productivity boost but nothing dramatic, and most features we're releasing are still human reviewed anyways.

Re: LLMs are eroding my software engineering career and I don't know what to do

#875

Post by camgunz »

Gareth321 wrote: post #863
camgunz wrote: post #833

Earlier quoted context omitted.

Nothing corroborated this. Performance on benchmarks has practically leveled off. The big gains have come from architecture (have a secondary LLM review output) or searching the internet. Also prices are going up. Everything points to the likelihood that we're at the top of the curve.

> Nothing corroborated this. Performance on benchmarks has practically leveled off. [There is plenty of data to support the claim that AI continues to improve, even exponentially.]( https://epoch.ai/trends ) As for benchmarks I feel compelled to remind you that as soon as a metric becomes a goal, it ceases to be a useful metric. The models optimise for solving the benchmark and we create new benchmarks to assess broa…

> [There is plenty of data to support the claim that AI continues to improve, even exponentially.](https://epoch.ai/trends)

This doesn't look at all exponential to me: https://epoch.ai/benchmarks?view=graph&tab=eci. OpenAI models went from 137 ECI to 159 ECI over about a year and a half, and the trends are similar for Anthropic and Google. These things have never been exponential.

> The models optimise for solving the benchmark and we create new benchmarks to assess broader intelligence. As models converge on 100%, progress obviously slows.

We are nowhere near 100% on important benchmarks like hallucinations: https://artificialanalysis.ai/evaluations/omniscience?model-...

...also, progress isn't improving with model releases.

---

We're running out of money. While we don't know how much it cost to train things like Claude, most (all?) industry reports indicate that a significant gain in function (2x) would require an exponential amount of resources (20x). No one's yet been able to convince investors that's worth it.

Also, we're running out of data: https://epoch.ai/publications/will-we-run-out-of-data-limits....

Also, we're running out of of low hanging fruit: "We find that the level of compute needed to achieve a given level of performance has halved roughly every 8 months, with a 95% confidence interval of 5 to 14 months. This represents extremely rapid progress, outpacing algorithmic progress in many other fields of computing and the 2-year doubling time of Moore’s Law that characterizes improvements in computing hardware (see Figure 2)." (https://epoch.ai/publications/algorithmic-progress-in-langua...). Maybe you think we'll continue along this breakneck pace, but again no investor thinks that, which is why prices are going up (investment is drying up).

Also we're running out of compute. Data center projects are stalling. Some of this is spiking energy prices, some of this is politics, much of this is grid constraints and supply chain problems: https://tech-insider.org/us-ai-data-center-delays-cancellati....

---

Finally, and perhaps worst of all, despite unprecedented investment data on the productivity gains is mixed. This is the biggest difference from other technological leaps like electricity, the industrial revolution, literally fire, etc. Those things were immediately, undeniably more productive. AI is not like that. You're not seeing an AI Microsoft, an AI Salesforce, an AI Oracle, an AI SAP, etc. You can argue that their advantages are structural, but there are no successful AI-powered alternative products (no AI Office, no AI ERP, no AI database, etc).

Re: LLMs are eroding my software engineering career and I don't know what to do

#876

Post by sankaritan »

This can surely feel painful, and I feel occasional sense of dread too, but then realize than not a lot have changed for folks working in medium+ size companies. Domain knowledge and tech expertise were never the sole differentiators for me when it comes to working with or even hiring capable software engineers. There's so much more to software engineer's job and ability to produce code / designs was barely 30% of our jobs. There is this interdisciplinary skillset we had to hone, how to work with others, how to understand what they actually want, how to ship things safely and learn from our mistakes and data to understand what to build next. You said you maintain candid relationship with PMs and stakeholders so you're likely underselling the whole scope of work you do day to day beyond immediate technical delivery.

The time ahead may be rough given the transition period we're in but Software Engineering role (or whatever we will call it in 1-2 years) should not go away.

LLMs surely will be able to do most of the individual work Software Engineer needs to do, but blending them all together is a lot harder task. And once we have AI doing that too, well, I believe at that point vast majority of knowledge work can be replaced by AI too and this is not a Software Engineering problem anymore.

Re: LLMs are eroding my software engineering career and I don't know what to do

#877

Post by onel »

pharos92 wrote: post #738
george_max wrote: post #617

I see many comments saying, "AI can't do X with 80-100% accuracy; therefore our professions are in good hands." While I don't want to sound overly pessimistic, the models are improving at a rapid rate. If asked ~3 years ago where the state of the models are today, it would sound like sci-fi if answered, "the models are creating full MVP apps in ~30 minutes with one prompt". The hurdles the models are facing now, like…

I would highly encourage you to watch this short clip https://youtu.be/5eqRuVp65eY?si=3fLT6S5q2OIUcu6r

That is a great video, but short is very subjective, the video is 24 minutes long :))

Re: LLMs are eroding my software engineering career and I don't know what to do

#878

Post by sensanaty »

george_max wrote: post #617

I see many comments saying, "AI can't do X with 80-100% accuracy; therefore our professions are in good hands." While I don't want to sound overly pessimistic, the models are improving at a rapid rate. If asked ~3 years ago where the state of the models are today, it would sound like sci-fi if answered, "the models are creating full MVP apps in ~30 minutes with one prompt". The hurdles the models are facing now, like…

Are the models all that much better? To me it seems the tooling surrounding the models got better, but the models themselves are basically interchangeable unless you're following a bunch of flawed overfitted benchmarks.

Also MVP apps are great and all, but I've seen 0 evidence of actually useful software from all this tooling, if anything all the software I'm using has just become more buggy and less reliable over time

Re: LLMs are eroding my software engineering career and I don't know what to do

#879

Post by mellosouls »

torben-friis wrote: post #625
mellosouls wrote: post #469

Earlier quoted context omitted.

LLMs routinely fail at our business specifics: Local tax regulations, particularities of the accounting process, specifics of our ledger implementations. This is domain expertise - software engineers are not needed for that. Ofc often senior sws are expert in it, but they aren't necessary . Traditionally its been useful for frictionless production to have engineers to be able to do maybe 90% of their work without con…

>This is domain expertise - software engineers are not needed for that. Ofc often senior sws are expert in it, but they aren't necessary. Our engineers frequently need to be on the loop with product and stakeholders: Due to real world messiness, many times the only true answer to "how does this currently work" is in the code. Enabling product and stakeholders to fetch that knowledge would be a giant time saver, so we…

"...place a non technical person..."

I'm not suggesting that and agree it would fail. Engineer expertise is important, but not in the old way.

Re: LLMs are eroding my software engineering career and I don't know what to do

#880

Post by treszkai »

jval43 wrote: post #831
davedx wrote: post #590

Earlier quoted context omitted.

Ha. I've worked in a fairly strongly regulated sector (energy, in the Netherlands), where I collaborated closely with our head of compliance, and she heavily over-interpreted the regulations while I often tried to find more pragmatic solutions. I think adherence to regulation and compliance is nothing to do with whether you're a SWE, a risk officer, or C-level, and everything to do with your own principles, ethics, p…

I've found two things to matter: 1. experience, i.e. knowing why and how a rule matters (in general, but also to auditors) 2. willingness to think If these aren't present, you get overly restrictive compliance that at the same time accomplishes nothing.

I'd also add a willingness to follow the (perceived) intent of the rules as opposed to gaming the letter of the rule. An experienced folk might say, "yeah it won't matter legally", and continue with "but it will matter to me because this rule is in place for this-and-that reason".
Post reply on HN