I would love examples of positions and industries where AI has been revolutionary. I have a friend at one of the largest consulting firms who has said it'd been a game-changer in terms of processing huge amounts of documentation over a short period of time. Whether or not that gives better results is another question, but I would love to hear more stories of AI actually making things better.
The category with the most dramatic gains for me: async business operations.
I run a set of AI agents autonomously — they handle email triage, content publishing, competitive monitoring, revenue tracking, and community engagement on a cron schedule. The productivity math isn't "I'm 2x faster at tasks." It's "work that used to require a person's attention 8 hours/day now runs in 2-hour cycles while I'm asleep."
Concrete examples from the past week: - 9 articles researched, written, and published without me writing a word - Stripe monitored every 2 hours automatically, escalating only on anomalies - Email triaged and drafted replies at 7 AM before I open my laptop - HN threads monitored for relevant discussions, expert responses drafted for my review
The key unlock wasn't a better model. It was persistent memory (so the agent accumulates context over weeks, not just within a session) and explicit escalation rules (so it knows the difference between "handle this" and "wake me up").
What hasn't improved: anything that requires judgment about people. High-stakes client conversations, pricing decisions where context matters, reading between the lines of a difficult email. The agent surfaces these for me — but I still own them.
The productivity boost is real, but it's better described as leverage than speed. The same person can now run more things in parallel, not do any single thing faster.
Three key things in coding which has helped recently:
1) Debug mode from Cursor. It highlights all possible hypotheses based on the steps to reproduce and whatever you know from the code. It slaps down logs and tests all these hypotheses at once.
2) Log reading, though this is a variant of the above. At times we get massive logs, like 10k lines for a bug. We go through what may possibly cause the bug - what screens, threads, race conditions, improper handling of ANY_KEY or default or something. Then we ask Opus to compare the logs to see if it matches the story. I ain't manually checking Line 314, Line 500, Line 44 to see if it shows in logs, but AI is just great at piecing together this whodunit.
3) Finding documentation. So there's been a few bugs unsolved around how very specific Android devices handle wifi reconnection or bluetooth or Ethernet. Instead of making assumptions, I can dig into the AOSP source code directly and see the conditions that trigger these. AI will even point out the exact lines.
I'm an engineer in the oil and gas industry. Some of my job involves messy judgement calls that I would never involve LLMs in, but some of my job boils down to integrating different items of data that exist across documents, drawings and a few databases that are in different formats and don't cross reference each other. At times I have used LLMs as a kind of "highly enhanced search engine" to do semantic search across documentation of every different types. My alternative was opening each document and using ctrl+f, along with my intuition of knowing what document titles to search for.
For a more concrete example, I have an interface to the data that comes from every sensor on the oil processing facility. It has a built in "AI" (I try not to use that term!) but it has a feature where I ask how to process data in plain language and it'll give me the calculations, then it'll also provide a plain language summary of all the calculations I conducted. That saved me 10 hours of work.
I am a negative nancy on LLMs in general but I still passionately believe that they're a tool which every white collar employee will need to learn to use effectively.
I cringe when I hear engineers say "I didn't know the answer so I asked ChatGPT" but I also do worry that I could be significantly outperformed by another engineer with 10 years less experience in engineering and 1 year more experience in judicious use of LLMs.
I find that I mostly use AI to compensate for Google search becoming unusable.
I use it more and more as advanced API documentation and writing code snippet. It made me a little bolder, and reduced friction from coding. I will test Claude Code in the upcoming weeks.
I work in cloud consulting + app dev. I’m always responsible for the full life cycle of a project - team lead, discovery sessions with the clients, working with them through the project and handoff.
I can now do everything by myself on most projects. Up to ones that would have taken at least 2-3 other people before. Before I would have had to delegate it just because I couldn’t do it all by myself on time. I know how to develop (professionally did 30 years) and I know cloud (professionally for 8 years). I just didn’t have time.
On the other hand, I haven’t done web dev in a decade. But I can vibe code an internal website for operations and authentication via Amazon Cognito. It’s just a free feature that I give them even when it’s not in the contract
Has it affected how you calculate/charge your rates in any way?
I don’t have a good comparison of what rates are like for independent consultants. I have only worked full time salaried for consulting departments - at AWS directly (Professional Services - full time blue badge RSU earning employee) and now for a third party AWS partner.
But let’s say I make more as a staff consultant considering my actual billed hours, benefits (401K, health insurance, paid time off etc) than I would make working independently and I always know how much I’m going to get paid.
But I will say in today’s cloud consulting environment, it is a race to the bottom unless you can lead projects and even then there are relatively few high paying (over $200K with benefits and 80% utilization not including PTO) outside of being a full time employee working in the consulting divisions of AWS, Google (they had an RTO mandate so I ignored recruiters) or maybe Oracle.
The only reason they can justify American rates for someone like me who knows cloud but who is mostly a developer is because I can lead projects and churn out code quickly - thanks to AI.
These "questions" just seem like a trap/disingenuous. The people "asking" commonly have decided the answer and are just looking to reinforce that by playing "No true Scotsman" with other people's productivity increases.
We have software written in what is essentially an obsolete platform (a RAD solution from the 1990s). We've been slowly hand rewriting it into React. A single developer converts one a week by hand, and the results are good (we still have over a hundred).
We've developed an internal tool to turn the software's definition files into a single XML, and then are feeding it through a multi-process Codex pipeline (multiple instruction files, that product intermediaries), which ultimately outputs a "90%" working React page.
Our developers then PR the output, fix/adjust/test, and release. We've gone from one per week for a single developer, to roughly three per week without a marked quality loss.
It is taking care of most of the repetitive parts of the work, and allowing our developers to spend more time in technology they're familiar with (React) instead of legacy stuff. Is it perfect, no, but the cost/benefit is clear.
As a founder, it helps me keep track of things by understanding our user session recordings and telling me where exactly they're going wrong.
I got same results in few minutes and cents that I got from a lawyer in few days and thousand dollars.
I imagine content farms, social media bots, at-scale scam operations, blackhat seo, busywork homework, social media propaganda botnetworks have all seen a significant boost of oroductivity.
Me! I'm in a leadership position at a marketing agency. I'm a self-taught programmer (just found it interesting and started on my TI-83 in high school doing BASIC and then some Z80 assembly) and studied creative writing in college.
I used to hand-write a lot of the scripts we used to automate processes (mostly Google Apps Scripts and Python). It was very helpful to us but slow going and sometimes frustrating.
In the past, I was limited by my programming knowledge and ability. Now I am limited by my knowledge of our business and ability to explain what I want to happen.
With LLMs, I can generally 1-shot most things I want to automate, including things that would've taken me days to figure out in the past. I generally just say that I'll spend 30 minutes on something and more often than not, it's either done or very close at the end of that time.
[dead]
[dead]