I could go on and on, but Claude recently decompiled the firmware of my camper van, documented all the CAN interfaces, then programmed an ESP32 module to talk to the van’s integrated systems (power, HVAC, lighting, tanks). That sort of embedded systems integration is completely out of my wheelhouse.
I honestly don’t understand AI naysayers. I use Claude every day both professionally as a Solution Architect and personally in a variety of projects I simply could not have ever approached alone.
> projects I simply could not have ever approached alone.
I think that's part of the divide between enthusiasts and naysayers. If you use GenAI on things that you couldn't approach alone, it's an incredible tool. If you use it on stuff that you're pretty good at, it's not a gamechanger (and if you're an expert, it's a minor boost at best). Many people's job are about doing what they're an expert at.
And in a team setting it can really accelerate tech debt especially if used by people that know just enough to be dangerous.
I find it's a huge boost for my day-to-day work.
If you work on architecture and Claude docs, then you can essentially just have it fill in the gaps. Work then mostly becomes a matter of defining what the next piece of functionality is (which you can also use Claude to help with).
The stuff that used to take days now takes hours. It's not perfect, but if you get your codebase into a good shape then the payoff is huge.
I re-read something I did 6 months ago doing this.
It's so obviously AI and had much less value than I thought now I look at it with fresh eyes.
Worse it doesn't read like I wrote it, I don't recognize myself in the doc.
I think part of it is we often notice bad AI usage. The llm generated "art" by someone with bad taste, or the patches to open source projects by people who cant program at all and are teerrible.
If the use is half decent people just dont notice it.
Anti-AI zealots (from a practical usability position. Not necessarily the moral ones) are like the people who looked at The Daily WTF and decided no humans are capable of programming. They had plenty of examples to point at, but refuse to look at decent to great programmers. The stories of "The AI deleted my database!" are prevalent and boosted by these folks because it confirms their biases. It literally doesn't matter if the LLM wrote strong warnings about the action about to be taken. They don't see that aspect of it. Just the fact that someone claims "The AI deleted my database!" is enough for them.
Despite all the liars telling me gaming is easier on Linux than Windows, most new games have some sort of issues launching with default settings. CC is able to dive into both the exact error logs and the recent community feedback on what tweaks / configurations are needed to make it work. I rarely have to go beyond two prompts before a game is playable. CC and Proton are enabling the Linux gaming experience far more than Linus ever has or ever was interested in.
> Despite all the liars telling me gaming is easier on Linux than Windows, most new games have some sort of issues launching with default settings. CC is able to dive into both the exact error logs and the recent community feedback on what tweaks / configurations are needed to make it work. I rarely have to go beyond two prompts before a game is playable. CC and Proton are enabling the Linux gaming experience far more than Linus ever has or ever was interested in.
Heh - I've just gone through a similar journey transitioning from Windows to Bazzite to play Steam games on Linux. I wouldn't have bothered pre-LLMs because my day job is Linux/Software and the thought of trying to fix issues here just to play games put me off.
While I think this is true
> If you use GenAI on things that you couldn't approach alone, it's an incredible tool.
I think this isn't true in all cases
> If you use it on stuff that you're pretty good at, it's not a gamechanger (and if you're an expert, it's a minor boost at best).
I think even then there's a divide.
I mostly work greenfield projects (and love it!). For these, AI has been a literal game changer. Our projects are built faster, with one or two orders of magnitude more automated tests, and all quality metrics are up.
Meanwhile, nearly all of my friends complain that AI doesn't help them. But they mostly work in very large existing codebases.
Still, even in large projects I think AI (the expensive variant) has been a complete gamechanger for me. Sure, I spend a lot on tokens, but I just feel happier and enjoy what I do more. The singalong people say about "thinking at a higher abstraction level" is what I feel. I really am thinking about architecture and larger patterns, instead of the boring nitty-gritty (which wasn't boring at all when I was a kid learning to code!...)
I think a key factor in all of this, to me, has been dictation. Most of the time, I don't write -- I use voice-to-text. I don't even read what comes out of it -- the LLMs get it (it is mostly unintelligible to anyone else) .
This means when I'm planning a big feature, I give a gigantic brain dump to the LLM in perfect stream of consciousness way, going through ideas, pros and cons, edge cases, what exists, what doesn't exist, where I'm sure of something, where I'm not sure and want the LLM to browse the state-of-the-art. Sometimes I spend 20 minutes just talking to the microphone before I send the first prompt. When I pair that with Opus, I find that I am able to build much faster and to go through alternative designs much more frequently as well.
I keep trying to tell all my friends: use voice to text and braindump to the computer. But they refuse... I couldn't imagine having to type everything nowadays. Even though I'm a fast typer, it's still much slower than the speed of my thought, which, granted, is still faster than the speed of my voice.
In effect, I filter much less, but I've come to think that's positive for the good LLMs: I throw all the edge cases and what ifs I'm thinking about -- all those years of experience dealing with similar systems.
If I wanted to go back to work in-office, that would be my major problem: I need to be able to talk with my computer all the time, loudly, and pacing through my room.
How do you use voice-to-text? You mean, in the browser? I am only familiar with Claude Code, which I have installed on remote server, and there obviously, voice-to-text does not work. I have to type, which is tiring.
I’ve installed Hex on os x. You just hold down a hot key to talk and it writes into whatever text entry widget is focussed.
There are many tools for this, and I use the one that I tried first, so there are probably better-suited alternatives out there.
I run MacWhisper, and I paired it with BetterTouchTool so it triggers on any input when I double tap the fn/globe icon.
Obviously all of my transcriptions through it are entirely local. I usually use the Large V3 Turbo model, though in the beginning I used Parakeet v3, which was slightly faster but produced more mistakes (and kept a lot of filler words -- 'ahhm', 'hummm').
However, if I'm interacting with the Claude or ChatGPT/Codex apps, I often use their voice recognition instead, because it tends to be more accurate, especially with punctuation, albeit significantly slower. OpenAI's is noticeably better than Anthropic but I feel like that gap has closed a bit recently (might be all in my head, though).
Like I said I don't really care about mistakes in the transcription. If you try to read it, it feels like a fever dream, but the LLMs get it.
If I say "taken" it may have "take and" If I say "all the while calling the method" it might have "although a while. while. call in the met of". This is a rather extreme example but I've seen them happen. The repetition of words happens because I'm talking with "humns and ahs" and do repeat words or just the ends of words. It's very rare for the models, especially Opus, to have any issue with this transcription. When they do, they tend to signal to me they didn't get it, or I catch them in the act. But, like I said, it really is very very rare.
As an example, I've got quite a significant feature to work on, which would have probably taken me weeks to design and implement, and I've used this exact method today to ink out the plan:
- I have spent the last couple of days researching the feature in my off-time and just "thinking about it in the background" (think: I fall asleep thinking of it -- a habit I've always had)
- I spent ~25 minutes brainstorming out loud. The transcript ended with ~17.000 characters and ~3.000 words.
- I sent that transcript, in cursor, to Opus 4.6-High with instructions on how to iterate on it and how I want to work while planning
- I then spent about 1.5 hours with it iterating and building the actual plan (and supporting technical decision document, which points at the FULL transcript of the whole interaction). Many of my original ideas made it to the final plan, others got scrapped or simplified, and others still got added. It contains a mixture of my ideas, Opus' ideas and our push-back on "each other".
- Now I have a multi-step plan, with at least 8 distinct stages to implement this massive feature which I know for a fact would have taken me weeks to implement, and I expect to implement it in at most 3 days, but very likely it will be a day and a half.
Final context (with regards to your Claude Code question): My main development environment is Cursor, though for personal projects I also use Codex and Claude code. For the initial "researching of the feature in my off-time" I often have interactions with ChatGPT and Claude where they have no access to the codebase, and I have them go find out what the state of the art on specific topics is. All of these interactions also involve me using my voice to talk to them (though nowadays I don't typically use their voice mode, I just let them reply in text). Then I brood over that.
Try using Handy! It's open source and free: https://handy.computer/ and also on Github https://github.com/cjpais/Handy
Yay for dictation! It's so nice to just think aloud and then have an easily editable record of your thoughts, even when you aren't feeding the outputs to LLMs.
This is exactly my workflow and it’s just incredible. I use aqua and wispr flow depending on which one seems to be returning the best results that day.
I'm about to complete a new non trivial functionality in a project of a costumer of mine. I spent an hour writing the spec. Then I asked Claude (Sonnet 4.6) to check if I missed something. I did, the sort of minor issues one notice after starting writing code, edge cases etc. That made me think about more issues and after a few iterations we settled down on a spec. I asked Claude to make an implementation plan and we ended up with 9 steps. It wrote the code for a step with new automatic tests and I performed some manual QA, which found further issues we didn't think about. We are at step 8 of 9 in about 12 hours of work. I would have needed a week to be there alone, with time spent researching and fixing bugs I created along the way, an inevitable part of our job but not exactly the most pleasant one.
This speedup is great. It improves the overall quality of the product (as perceived by the users) because I can ask Claude to add features that my customers and I would have dismissed because they take too long to implement. We would have settled down with a more basic UX.
So is it a game changer? It is in the same way those HTML / CSS framework like Bootstrap were game changers: suddenly every developer could create a decent and consistent UI in a fraction of the time with a few bells and whistles that we wouldn't have bothered coding. As a side effect a lot of web apps felt look alike mass products and web designers had to reinvent themselves, but the economics leaded inevitably in that direction. Would I spend again one of two weeks doing alone what I could write in a day or two with a LLM? Not anymore, not at this cost ($20 per month.)
I'd love to read a full transcript of someone going through this kind of collaborative programming. I see this kind of process mentioned a lot but can't quite figure out the details in my head. If anyone has a link to a blog post or similar showing this process in depth, I'd love to give it a read :)
Jon Gjengset has some live streams where he does agentic coding.
I think it will click once you actually sit down with the AI agent, toggle Plan mode, and just tell it what you want to do in couple of sentences. It will immediately start building up the plan, presenting it to you what it thinks is the right approach , with the steps to take, with open questions that you can look at and answers. Then send them back to the AI. Repeat. That process along would give you a progress way further than you try to do it by yourself.
You can tell it to start implementing step 1. And you pick it up from there. Very natural how you would approach an expert for help, but you can always audit.
I did not use plan mode but I'll give it a try.
I can't provide a transcript because it's work I made for a customer and I'm bound to a clause of confidentiality.
What I did is what I use to do while starting to work on a major feature: make a list of changes, new and modified functionality, think which code and db tables I will touch and how, set constraints on the edits (eg: that API must not change, that one must be retro compatible) etc. I've been a bit pedantic because this time I had to tell it to someone else. I wrote it into a md file and asked Claude to check the code and find out if my plan was consistent with the code we were starting from. It made a list of things that I needed to detail more, added some questions and we iterated on it. Basically it's what I do myself but it happened faster.
I've been using superpowers [1] for this purpose, and have really appreciated how it guides the model to use careful, methodical approaches to answering my prompts. It's great for multi-step planning, design, and implementation, but also has guidance for debugging, accepting a code review, etc.
[1] https://github.com/obra/superpowers
Yeah I feel like I’ve learned a lot from superpowers. It’s such a thoughtfully developed skillset.
This Spec Driven Development Short course on DeepLearning by Paul Everitt is a nice 2-Hour walkthrough: https://www.deeplearning.ai/courses/spec-driven-development-...
You describe almost exactly how I work, except I always use Opus with effort locked at max. Lots of detailed multi level planning, then coding the different planned steps, which it at that point just one shots, with a plan review and adaptation after each step.
5x speedup and quality.
The dangerous thing is when you’re a novice and can’t identify the BS. That’s why for people with “good” and “expert” skill, it’s not a huge boost. They can identify the BS, and what’s left is modestly helpful.
The highest danger in using AI comes precisely to people who stand the most to gain from it.
Exactly that. Novice don’t notice the BS. But they see the output and it looks magical. The UI is working! Hardly any time to code that in
Then they send that PR for a review by a more senior person. And that senior person doesn’t even know where to start on how to explain why it’s all wrong and likely to collapse in prod.
Tons of good use of AI. But tons of bad use of it. And when it’s bad and people don’t notice it, that gets dangerous. So because of that, now we spend a lot more time in doing reviews. Essentially creating a new bottle neck
> If you use it on stuff that you’re pretty good at, it’s not a gamechanger (and if you’re an expert, it’s a minor boost at best).
This was probably true last year, and it’s a common talking point, but I’ve seen too many examples now of deep experts using Claude & Codex in the last year to solve very big problems, and write or rewrite large systems. The experts do complain that the LLMs can sometimes get stuck or go off the rails and they need to pay attention and actively steer. But nobody I know who’s using it is still claiming the LLMs aren’t a game changer, even quite a few people who were staunch holdouts for a long time. I was skeptical myself, for a long time, but had my oh shit moment late last year.
One caveat - to get expert results, you do need to have some experience using LLMs, you need to use it to write plans and design docs, know how to use ‘skills’ and MCPs, use it to review code, and (for now) you need to understand context compaction and when/why to use sub-agents. If you’re a domain expert but an AI noob, it’s less effective than an expert who knows how to use AI and has experience.
One of the biggest problem with humans is we’re wired to spot patterns and draw conclusions and then we have a really hard time seeing and accepting change and updating our mental rules. The LLMs are getting better. They have already gotten better, and they’re going to continue getting better. It’s too early to draw conclusions, and many conclusions people have already declared are out of date and no longer true.
Imo it's still great for areas you have expertise in, because it's a tool for automating the boring, repetitive, or time-consuming bits that you can then expert-verify.
I'd rather review & tweak generated test cases than write a load of boilerplate, test setup, etc. myself.
Classic Gel-Mann amnesia...
Using GenAI on things that you couldn't approach is also extremely scary and dangerous in my opinion. For example, I would never in a million years use generated code I don't understand (fully) to interface and possibly interact with a physical object that can fail, catch fire, break etc. in case of a bug or misuse, like OP mentioned.
I'm an expert at datalakes. I manage them for my company. I also am proficient at backend web. Even still I use Gen AI frequently to manage it all. When my company downsized I kept the lights on. Not enough bandwidth to do more. I've since materially improved the system and doing things we never did even at a team of 2 or 3.
Outside my expertise I've begun writing static recompilers for old retro game systems and have gotten some games off the ground. I understand WHAT they're doing but I neve had the expertise to do such things myself. Even if I did I could never operate at the velocity I am now.
Same. I'm a DevOps engineer, so a jack of all trades master of none type of guy, and Claude Code backfills my knowledge gaps and turns me into kind of a superhero. I think it's key to already have a pretty good idea of what you're looking at, though.
>I could go on and on, but Claude recently decompiled the firmware of my camper van, documented all the CAN interfaces, then programmed an ESP32 module to talk to the van’s integrated systems (power, HVAC, lighting, tanks). That sort of embedded systems integration is completely out of my wheelhouse.
Can you tell me a bit more about the firmware/camper van? Has this firmware been decompiled by anyone else?
I am more of a "huh, interesting demo, I'm gonna check in on it later" sayer than a naysayer. My biggest reason, with coding, is that I already, before AI, struggled to deal with too many distractions from my coding and too many piles of low quality output. I should probably check in since it's been a bit but every time I've tried to generate some simple project, I look through it and think what terrible garbage with so many errors. After two decades of developing my craft, I struggled with most of my fellow human programmers too. The business loves delivery it now even if then someone is revisiting it hundreds of times more to fix it in little bits for a total effort cost of 10-100 (or higher) times more.
> I honestly don’t understand AI naysayers.
As an AI naysayer, I see and appreciate the productivity gains, I don’t like the associated cost, mostly the spike in workflow centralization and opaqueness.
Yes and I'd like to talk about the environmental impact as well please.
I can do things at least 10 times as fast as coding myself now. I'm pretty sure the environmental impact is, frankly, a reduction. My home computer (and feeding myself) for 10 times the time is more than I'm burning in a data center using an LLM.
Indeed.
I often see comments on the ethics and water/energy consumption of AI, but very few for non-vegan* developers boiling water for their coffees in air conditioned offices that require a commute, which I assume is a common default.
*blue-water usage of a burger patty is worth looking up, I was astonished
I assume you mean car commute, right?
I don't know about the US but in France (or Europe) the environmental impact of air conditioning, car driving and meat consumption is well understood by most educated people.
I'm not sure I understand. You realise that our token consumption in coming on top of our other energy/water consumption right?
Also, the environmental impact might be small if we run a small number of inference queries every day, but that impact will grow linearly with the number of queries and agents we run.
If I spend less time working, which has been the big outcome here so far, I get more life. My own energy usage when I'm not working is mine, not part of the work.
A lot of the time people relate an anecdote about how Claude helped do some cool thing, my reaction is that it's not a thing I would have thought about doing in the first place, and that I still can't really imagine wanting to do myself, even though it indeed sounds cool.
This is no exception.
You will be surprised that there are lots of things you want to do yourself but haven't been able to (not just ability, but time and effort).
> I honestly don’t understand AI naysayers.
If you are describing someone that wants things to be done, then I agree with you.
If you are describing someone that wants to learn things and do the fixing themselves, then I don't understand how you could say that.
For a lot of us, the learning and the mistakes and the eventual fixing of a thing or completion of a project is the goal. Us doing the work is the reward function. AI strips that off and simply finishes the project, removing any and all incentive for the person involved, if they are this kind of person.
Again, simply having the effort completed is probably the goal if you simply want to have something completed that was not completed previously, but if you are someone that derives satisfaction or dopamine from doing the work yourself, then it is very clear that AI completely short-circuits this reward path for that person. Those are the people who don't like AI, and they have a very solid footing with that argument, I think.
I know Anthropic has blocks on using Claude for security reasearch; Are they not blocking Reverse engineering or RE tools?
From my experience, the safeguards only come up during exploit development. You are free to do reverse engineering and even the first half of vulnerability research (i.e. vulnerability discovery) and it only stops once you want it to actually write the exploit.
My opinion is that it’s a defensive mechanism. I’ve seen it across experts in knowledge domains and my own. When you hear experts disagree it’s fine because it’s another human, when the LLM disagrees and provides an objective backing that’s often solid, people jump to defense and look for very subtle nuances they wouldn’t bring up with peers and those subtleties are often highly subjective and arguably often incorrect. That’s been my observation.
I for one welcome our new LLM overlords so long as some provide be solid living standards. Mistakes do happen and they aren’t perfect so experts often do have arguments but they do come stupidly close to approximation of expertise.
Maybe because the scale of investment out strips the value?
What trillion dollar problem is AI solving?
If you're going to put it that way, companies, globally, spend something on the order of $20 trillion on office workers. If corporations didn't have to spend that money on them, and everything else in order to support them, they wouldn't.
Then the workers wouldn't spend 20 trillion and the economy as a whole would tank.
Because work just became unfun. You are context provider to LLM.
What is exactly your work? Give context to llm, review, update context. Navigating some sort of super intelligence thru your company's harness is not the same as writing code and creating ideas from scratch. And I don't understand what's fun in that
Yes, you can ship software faster, make corporation even more money. Why is this even important for regular worker? I liked the craft itself
Any docs/suggestions on how to do this?
Would love to do it with my campervan as well :)
I'm interested to hear how you use it as a SA
> I honestly don’t understand AI naysayers. I use Claude every day both professionally as a Solution Architect and personally in a variety of projects I simply could not have ever approached alone.
My brother works in wildlife trapping and management. I've been brainstorming and prototyping ESP32 sensors and mechanics for traps and educational devices with him. I probably won't end up doing the work with him, but I want him to see what's possible with my other brother, a machine learning expert.
Nothing has been deployed in the field.
Nothing will be until he and my other brother commit and get proper software risk management policies in place. And legal advice and other support work. And honestly, he's been careful and hasn't pushed for deployment.
He works with rabies. He works in people's neighborhoods. Maybe yours. Do you want me to finally get a Claude Code account created and go wild building shit, or keep reading up on ISA 62443 and other security frameworks and mapping out the risks?
I'm not going to drag LLM generated work into your neighborhood today. Would you honestly want someone else to?
And when people realize this is happening now everywhere and the entire AI industry is fucked, including other machine learning fields that get hit by association?
Then my other brother the fucking Princeton Machine Learning Super Star can't pay for his fucking kids' schooling because of a million people fucking *not understanding* and intelligence agencies taking advantage of it. He's smeared by the broad anti-AI brush.
Then my brother may have to depend on more assistance from law enforcement, legal resources and conservation agencies. Because I didn't have the power to stop the LLM hype machine earlier.
It always would have made sense to have them work with state and federal Wildlife Conservation officers and agencies. Now it feels a little less like watching my brothers build those relationships out of mutual respect for other professionals and more out of need. It feels unequal.
So, I have to put in work today assuring my brothers' clients of tomorrow, who care about their family and kids, that no, their machine learning algorithms won't take their elderly parents medicine and push them down the stairs. It will, with careful review from lawyers and experts, help their kids identify nature in their backyard on their Smart TV. If they want, it will identify the difference between gopher tortoise holes and mole holes, and maybe if they opt-in to a Community of Saving feature, it will let Fish and Wildlife Conservation know there are habitats nearby so we can see how healthy our ecosystem is as a community together, or call their preferred pest eliminator.
That sounds like PR. Because I have to do that extra work today. Because otherwise we aren't just protected by our fellow professionals who care about theie work out there in the field, instead we always need some bigger institution protecting and controlling us.
My brothers are delaying committing resources to projects. That's fine, they have other important priorities, but I keep warning them. And there probably will be an equivalent of the "Video Game Crash of 1983" in 2028 or whatever. And I think if I had had more personal power and been believed I could have done something about that before we had to be protected.
I'm sorry, I said I probably won't work with them. That shouldn't be a probably, it's a won't work with them.
I hope they are inspired. I hope others besides me can advocate for UBI which would have lowered the pressure for irresponsible AI use. I hope I have easy healthcare options with a doctor without making it my adovcacy or story.
>projects I simply could not have ever approached alone.
Learned helplessness.
That's not fair. It often has to do with limits of time and energy. There are countless things one would do if it took a few hours, but which one doesn't have a few days to spend on.