I have a similar claude story (much less money though), with the IRS R&D tax credit. The auditing firm initially said we qualify for $0. But then I had claude analyze past R&D reports and our expenses and it found the problem. The auditor had miscategorized our company.
So claude drafted an email even pointing to the right Internal Revenue Code (IRS Law), and specify why we fall under a specific category. The auditor got back to me two days later admitting their mistake and said our company now qualifies for $8k in tax credits. And a few months ago, it identified items in our AWS that saved us $250 a month (paying for itself).
So now I joke that even if I have a claude max plan, I've still come out ahead financially.
By my estimation (guess) you won't actually need to spend that much because the models are already getting a point where they don't need to get a whole lot better to be extremely helpful across many domains.
And it looks like those very helpful capabilities will continue to transfer to smaller models as well, as architectures and training regimes continue to refine.
I can fairly easily imagine a world where the only people needing to spend a lot of money on models are those that are using them to solve truly novel problems. The rest of us will get plenty of use at reasonable costs for the typical day-to-day helpful stuff.
It highly depends on the task. For math and coding, sure. But for knowledge tasks
GPT-4 is wayy better than even SOTA ~100B models. For my knowledge test cases the lines get blurry at >400B
There seems to be a mass delusion about how capable SOTA models actually are. That's my only explanation for how poorly I find them performing in basic knowledge tasks compared to how others describe their prowess.
I understand you to be implying that I shouldn't trust my perception that there's a meaningful difference in how much different models hallucinate. I will take that under advisement, but I am still interested in the answer to my original question.
I am eagerly awaiting being able to run a strong local model. I'd hand Apple $5k right now for a Claude in a box. I know the cost might not be there now, just saying that is around my ideal price point.
$10k might even be worth it - but i'm assuming that the more expensive it is the beefier it is too, which also means more electricity.. and i already run ~6 computers/servers in my house. If a power surge happens i'm going to go live in the woods lol.
I would do the same but my issue is that the models are changing so fast, so I don't want to be left out of the next model cuz it only runs on an even newer GPU or something like that.
But maybe my limited understanding is thinking of this wrong.
I've run the latest local models over the last year, including the recent Qwen 3.6 30B A3B, on a 9yo GTX 1080 and 32G RAM I have lying around[0]. If I can do that I don't think hardware will be a problem for you in the near term. The only updates I've needed were to Llama.cpp when a new class of model was released.
[0]: In my case, I want to see how local models perform on limited hardware, sacrificing context size and intelligence compared to SOTA models, so I have to really limit my expectations.
> I would do the same but my issue is that the models are changing so fast, so I don't want to be left out of the next model cuz it only runs on an even newer GPU or something like that.
I think the same, and it's why i stopped caring about running llama/etc at home last year. That coupled with the models being dumb by comparison to SOTA really make me fine with waiting.
But in a year or two it's going to be difficult to resist at home, assuming the pace of improvement holds.
I do, though they're not as bullet proof as you'd hope to my understanding. Hell i have one at the house level too - since i have an EV sitting behind that as well.
In my anecdotal experience there is a huge gap between GPT-5-mini which hallucinates relentlessly and Claude Opus or the latest GPTs which are fairly reliable. I'm hoping that gap can be closed with improved approaches for small models and that good reliability is achievable for LLMs without requiring absolutely mammoth computing resources.
For what it's worth, I also used GPT-5.2 (via duck.ai) this year for questions about taxes and it was helpful — which makes sense because there's an abundance of material about taxes out there to be synthesized, so a text predictor trained in that domain should do well.
[sci-fi “AGI” scenario] What if those with elite model access philosophize in a way us mere mortals can’t understand, so the elites have to prechew the ideas for us to bring them to our level, and they control the narrative?
In reality now, curious about social implications generally. Does this go beyond problem solving? Maybe the intelligence per token you get via your free library card/membership is insufficient to compete with peers in dating/employment/etc. markets, thus puts you at disadvantage.
That isn't really philosophy, but rather doom and gloom theories.
Control the narrative on what exactly, how I write a bootstrap script for my servers? Or what type of flower is in this photo. Not everything is politics luckily.
To be fair regarding taxes it should be that you get all exemptions by default and the other side telling/justifying you why you don't qualify, instead of you not getting the things you should because you're not sure how to interpret the law or don't know this or that rule exists. Taxes shouldn't be that difficult, and the US version of it seems to be behind even the one I have here in europe (where my taxes are done "for me" for anything non business owner related). If the government is calculating your taxes anyway, they should just give the number to you instead of asking you a number and you better have the same as us or you're guilty of something.
Feels like a system that is deliberatly made to be more punishing for those who can't afford the help or the education to figure things out.
I would agree with you on most situations (like 1040 personal income taxes especially).
But in this particular tax credit, there's no way for the gov to know automatically what percentage of payroll was spent in qualified R&D expenses, since it's day to day business operations. Which is why we are _forced_ to hire an outside firm and pay them thousands of dollars (when Claude did an even better job), just to analyze how much of our time qualified as R&D expenses.
The problem I have is that I am forced to have to find a firm to do this, and most firms won't even work with companies as small as ours. So then we're stuck and losing out on years of R&D tax credits at the moment, when I really don't need them anymore, to be honest.
> it identified items in our AWS that saved us $250 a month
I gave Claude Code access to the Azure CLI and had it review usage. It found a couple servers that were underpowered, but it also found a lot of things that weren't being used at all. The total reduction was $40k a year.
Kind of. I first of all did the entire return with it. So we went step by step and yes I fed the forms one at a time. So I filled in 1040 as best I could. Then just asked it what to do next at each step. It helps I've done it before so most of the steps it returned were ones I've done before. However, it did mention several things that I had not heard of, and also some new taxes that I had to file due to some exceptional events last year. So all in all, a solid use case. This year I have an accountant, but it saved my butt this last year, and I will absolutely run through my accountants decisions with it. It has an encyclopedic knowledge and an immense capability to understand without getting tired.
Well my tax situation changed a bit this year, so there were some new taxes I had to file. Additionally it helped with depreciation for real estate. Helped me identify the exact part on the appraisal where I would get the cost of the dwelling (vs the land), and then walked me through the various ways the IRS allows people to do this. Helped me do some tax planning for next year as well around mega backdoor roths. Not lifechanging, but useful.
What's the threat model here? OpenAI gets my social security number and Sam Altman steals my identity? OpenAI leaves an S3 bucket open to the public and my filled out 1040.pdf gets leaked to the world?
Oh no, OpenAI knows how much money I make and they're going to send me ads! Ads that are relevant to my interests. How connivingly evil of them!
People's tax returns are essentially public (yes I know they're not allowed to disclose them). Didn't send the forms in with social security numbers.
This absurd concern for privacy is silly in my opinion. The moment something is submitted to the government it ought to be considered public. Even your social security number is essentially public for anyone who cares to find it.
I would not submit my bank account information to these services, or my passwords, obviously.
Honestly, tax returns should be public again. Would make everyone better behaved IMO. It was this way for most of American income tax history believe it or not.
To be clear, my information has already been part of several breaches anyway. What protects you ultimately is the law not information security. Of course this point is often lost on engineering / computer scientist types who don't understand how law works.
I use AI to help me do my accounting (how to categorize and account for things). It pays for itself because I need to spend money for far fewer hours from an accounting firm each month to make sure that captialized expenses, depreciation, tax credits (we have Historical Tax Credits from restoring an 1880s building) etc, is put in the books properly. The AI gets it right a high enough percentage of the time that I only the to have a real accountant look at things once a quarter to make sure it's all OK. I used to have a dozen questions every month.
Claude Code is really good at stuff like this. The other day I tried to recover some images from an SD card that had gone bad. I used GetDataBack to recover files, but they appeared to be malformed and didn't open in image viewers.
I tasked Claude to analyze the files and figure out what's going on, and eventually we figured out that each file had a custom metadata header + thumbnail + actual image concatenated. I had it write a python script and was able to recover all the images with their metadata. It's nothing a human couldn't have figured out, but it was definitely WAY faster than doing it myself.
I've also used Claude in the past to figure out how to break into routers with locked down firmware. It's great at suggesting and trying different approaches.
I have a friend that just picked up a new consulting job resurrecting an ancient Windows desktop application. No source control, no tests. And it's spread out over a dozen different folders with names like "_old", "_new" and "dates". Claude's doing a tremendous job in getting him to grips with what is actually happening in the application, what's relevant, what's not, what's different. I think it's literally saving him days and days at work.
if your friend has access to the binary and can pull it out to different box, they might get a lot out of a ghidra mcp -> https://github.com/LaurieWired/GhidraMCP
I'm not well versed at reverse engineering binaries or interpreting C/assembly so ghidra MCP has been an absolute gamechanger for helping me write tools. Once my project is complete, I plan to learn how to do the analysis myself manually and have cc guide me along the way.
> I have a friend that just picked up a new consulting job resurrecting an ancient Windows desktop application. No source control, no tests. And it's spread out over a dozen different folders with names like "_old", "_new" and "dates".
That doesn't sound very impressive. Not being tracked with a version control system is fixed instantly with a git init, git add ., git commit .no AI required.
Covering the app with tests is also something that requires no AI. At most, coding agents can generate characterization tests in broad sweeps, but we are talking about a delta between hand rolling and vibe-coding of a couple of days.
Where LLM shines is helping developers build up an understanding of what is in place. Running /explain on a codebase can quickly provide you with a high level summary of what's in place.
The relevancy here is that he's denied the git history, versioning, branches, implicit documentation that even bad source control practices would have given him.
That's what the comment is saying. In normal repositories, version control acts as a record of the momentum of the direction the product was taking. If it's just "_old" and "_new," the developer has to read and understand both, which I think is going to be far more time consuming than your estimation.
I'm sure data recovery companies are pretty pissed that slightly esoteric data recovery abilities are becoming more accessible for average software devs. They were charging an arm and a leg to remote in and run scripts.
They still have two important moats: (1) expensive hardware tools (even stuff like SATA write blockers are kind of expensive for what they are), spare hard drive collections to swap failed PCBs, etc and (2) the "nobody got fired for hiring us" edge similar to how everyone calls in Crowdstrike/Mandiant after an incident. If a suit-level manager finds out customer data was lost, they are going to want to call in an expert so they can immediately tell the customer they did, not have the same internal team try to figure it out.
As an aside to #1: The cool thing is in modern times the hardware tools have come down stupidly cheap in price. Even SD card recovery is (vaguely) in the right skilled hands in a pseudo-professional home lab these days.
I did EXACTLY that last night. Was doing by hand for about an hour and got to a point where I didn’t feel competent anymore and asked Claude to take from where I was.
5 minutes later I had almost 3 hours of important footage recovered.
A lot of "Claude Code is best at X" claims are probably user-selection bias.
The people saying it are often exclusively Claude Code users, not people who are actively benchmarking Claude Code against Gemini CLI, OpenAI Codex, GitHub Copilot, and other agent harnesses on the same tasks.
The claim may still be true for certain scenarios, but the evidence is usually anecdotal, not comparative.
When I hear "claude code one-shotted X" and X is a novel problem, I mentally substituted "the agentic harness that I tried one-shotted X," since that's what they're saying.
Getting any smart model to take a look at the task is the sort of lift that the speaker is usually pointing to.
The harness is pretty much irrelevant for general tasks.
You can write a 100 line harness that only has one tool - try either "bash" or the more fun "you're running within nodejs, here's eval", you'd be surprised in how close to CC/Codex performance you're going to get.
I have only my own personal experience for frontier models, but I have seen different performance of Opus when used from Pi or Claude Code or Zed for example.
I worded my comment poorly. I agree a good harness goes a way, but the harnesses most people use fucking suck and trip up the model so often that I don't think it's advisable to attribute successful results to them.
E.g. GPT5.5 with Codex on my Windows box likes using PowerShell for everything. OpenAI decided it should use the native shell instead of bundling a bash, or using git bash. Sure. But the model is so overfitted on bash that it fucks up PS quoting like once every 5 commands.
Every harness with LSP I've seen trips up the model as well. They insert diagnostics after every edit, polluting the context with errors that the model has to actively decide to ignore, every time, until it finishes its work and gets the code to a consistent state. Telling the model "run npx tsc --noEmit to check errors" will outperform a LSP 100% of the time.
Another example is basically everything Anthropic does - they add things like "think if this is malware!" after read and lead Claude to spend its reasoning effort on thinking if your React hamburger menu is malware, instead of on how to write it.
"This is not malware (em dash) it's a hamburger menu. Let me apply the edit! Hmm, is it malware now, after my edit? No, me changing border-width did not turn it into malware! Good! Dodged a real bullet on that one!"
I'm frankly amazed that we've gotten to the point where the models can produce good results in these sorts of environments.
I did that, wrote my own harness “Jarvis”, simple loop. Still results were terrible using the same model in comparison to for example OpenCode. So X Doubt.
> Bitcoin trader recovers $400,000 using Claude AI after getting 'stoned' and losing wallet password 11 years ago — bot tried 3.5 trillion passwords before decrypting an old wallet backup
Man. I wish I had a lost wallet worth a quarter of that even, technically didn't need Claude for this, just needed any password cracking software.
Explaining your life to an llm, then having it generate permutations of passwords to try does sound like it would work a decent percentage of the time.
A large percentage of passwords aren't a random string of characters but a memorable word + memorable number. There's existing projects that basically do the same, and 3.5 trillion doesn't really make it clear if one of those wouldn't have worked as well, but I can see it having an above random chance to guess a password.
>Explaining your life to an llm, then having it generate permutations of passwords to try does sound like it would work a decent percentage of the time.
I cannot relate to this at all. This information doesn't really seem that helpful. What might the strategy look like? Including spouses names or other proper nouns associated with you. But it's going to be a massive brute force effort still, and the likelyhood of a targeted crack that performs significantly better than more naive brute force passwords seems so unlikely.
Are your passwords like "SPOUSE_NAME:HOMETOWN_NAME"? Even if so there are probably more people with dictionary words that can be brute forced faster. IT would have to be the case that more people use patterns like that compared to something a regular dictionary attack could crack.
The amount of times I've gotten told a password and it contains birth year or anniversary year, maybe child birth year, is insane. I'd say 9 times out of 10 it's that or a dictionary word.
The idea that someone (the NSA?) is training models on all of our collected info, and using that to predict all of our hidden information, is horrifying.
The best time to start using a password manager was 10 years ago. The second best time is now.
I had a high school friend that died about 10 years ago from an over dose. He was always tech forward and had talked in the past about getting drugs from the dark web to sell locally.
I wasn't particularly close with him after high school, but he was an only child, and I can only imagine his (older) parents just tossed his computer. I wouldn't be surprised if he had had a few hundred BTC on there.
I have a lost wallet with about 300 Bitcoin sitting in a landfill somewhere. I tried out Bitcoin really early on and mined those over a few weeks. But they were worthless back then and I was burning electricity for "nothing" so I stopped. This was before that 10k Bitcoin pizza purchase happened. I have some regrets lol.
Way back in the day when Bitcoin first came about, I once idly contemplated spending some time and money on it just because it was a very cool technology. At the time it was a bit of a hassle because you had to mine your own.
Then I was especially tempted years later after running into the MtGox booth at CES, and seeing how convenient it had become. I remember asking a guy at the booth if Satoshi was really still anonymous or if any insiders knew about him, and he said "no" but was bit surprised I knew about Satoshi. I guess Bitcoin was still quite niche then even amongst a technical crowd.
I considered buying a few bucks worth of bitcoin then for lulz, but I thought that money was better spent on beer lol.
I've never really regretted spending that money on beer rather than bitcoin, because I knew that even if I did, it would 100% have been on MtGox and I would have lost it in the hack anyway, which would have been even more bitterly frustrating.
Yup. I was really close to buying $1000 (at $1) worth of BTC ages ago. I don't dwell because the stress of managing that through time would have eaten me away lol.
With that said, i do regret not at least mining/etc. Back then i could have mined in many ways, and getting into it as a hobby probably would have meant holding larger amounts of BTC in the long run.
I remember thinking about buying $100 (at $10). And then realizing I didn't actually know how to do it and didn't feel like looking it up or going through whatever steps to do that kind of transaction online, or worrying about getting scammed....
A friend lost £2000 worth of BTC in MtGox which is probably worth a fortune at today's prices. The last time I spoke with him he said there was some sort of lawsuit for victim compensation. How did you recover your funds?
Someone gifted me 87 bitcoins back when they were worth ~0. They are still in some wallet somewhere I guess, and I saved the password on a harddrive I threw out around the same time
I was making a long edit in a crappy wiki UI and my browser froze. It would have taken a long time to redo, hours.
I didn't want to take the chance of force closing and losing everything. I used claude code to extract my text out of the browser internals and filesystem objects.
Many crypto wallets use a key derivation function (KDF) to add an amount of computation (and memory usage) per password tried - to mitigate brute force of weak passwords.
The increase in compute (decrease in brute-force cost) combined with price increases in many crypto tokens means brute-forcing old wallets can become worth it years after passwords were forgotten.
And of course even smaller, local AI models can now easily write optimized scripts to brute-force any given KDF function.
The compute power needed use to be of the order of 5s per password try.
So it effectively mitigate brute force back them, you need a absurd compute power to crack them.
Moore law did its thing, now you can do it with a lot less computer power.
A wallet is just private keys of some specific public keys on the blockchain that have unspent output (UTXO). None of what’s described in this article involves the blockchain, only the storage and protection of the private keys on a local computer.
You can imagine that in your example, you didn’t change the locks on a house, but rather you put the house keys in a secure lock box and you changed the locks on this box.
Changing the locks on a house in this case means transferring from an old wallet to a new wallet and then abandoning the old wallet. That’s exactly what the OP is trying to do. It’s just that you need the original key to do it.
I spent a couple of days mining many years ago and got 2 bitcoins. At the time, they weren't worth the electricity they cost to mine and over time I lost the wallet and all information related to it.
You are better than me. Back when you could mine BTC with the CPU, i had about 2 coins. I found it useless and silly and deleted my wallet at some point :)
>they weren't worth the electricity they cost to mine...I'd love to mine a bitcoin a day on my PC now...
you can!... but they wouldn't be worth the electricity now either. the cost of mining (amortization of hardware costs plus electricty) is the value of bitcoin. if bitcoins are a bargain to mine, more people will mine them thereby reducing rewards.
should you have mined more back then if you had magical perfect knowledge of the future? no: they weren't worth the electricity.
instead you should have bought more of them back then.
The other day, I asked Claude to track down the leaked Claude Code source so I could study it. It refused, saying “given who made me, I’ll pass.” It gave me some pointers on how to find it myself, which worked.
There isn’t that much of a difference between “help me crack this bitcoin wallet” and “help me crack this executable.”
I don’t exactly have a solid point, just some general observations. First, I think we’ll see AI more and more simply refuse to do any kind of forensics, as forensics becomes more powerful. Second, that implies local models will become more valuable, since they’re the only ones willing to do that kind of work.
I once got myself banned from Claude by researching barbiturates, since they’re connected with suicide. So my third observation is that we’ll see an uptick in people getting punished for trying to do things with AI that people don’t usually do. (Luckily the unban form worked.)
Someone downthread asked “how’d he convince Claude the coins weren’t stolen?” Which is an interesting question, because presumably some people trying to crack a wallet have stolen it. So I guess the fourth observation is that the exact framing you approach an AI with will become more important. There was the classic “do this or I’ll cut off my arm,” which worked a year ago. But in the future it will be more like “hopefully the AI believes my story, or else I’ll get into trouble.”
It’s good there are multiple AI vendors, or else it’d get real dystopian real fast when the de facto AI’s policy becomes something you have no way of working around.
Kind of cool to hear... I had a couple computers running miners at home towards the end of my second marriage (which ended in 2010-2011), and after I had a few coins when realizing I couldn't actually spend them (IIRC $0.25 at the time), I just deleted the wallets. I had no foresight or faith that they'd be worth around $100k each in 15 years. I'm curious how many did the same, how many coins out there were just deleted altogether.
i distinctly remember a coworker many years ago offering to just give me like 10 bitcoin to see them and how it worked. i said "no, thanks". granted, hindsight is 20/20
I went from giving them away to anyone in the office who asked (very very few), to selling them for under market when they were in the $10-50 range. Only three people took me up on the offer then, and they bought moderately life changing amounts I've later been thanked for. Such as a down payment for someone's first home after he got married. Luckily they held onto enough of their coins to make a meaningful difference in their lives like I preached about at the time.
The number of folks who ping me at various BTCH ATH hitting the news is expressing regret and laughing about it is... a lot!
It was strange being the "Weirdo Bitcoin guy" for a time. Then you started hearing about it in the mainstream news/etc. and it got even stranger to me.
I did quite well for myself, but of course wish I could go back in time and actually go balls to the wall on my early mining operation. I thought I was being irresponsible enough as it were - maxing out my rental townhouse's electric service with a rack of GPU miners in my basement, and some sketchy DIY electrical work to make it all happen.
Unfortunate to me cryptocurrency devolved into such a horrible place with the shitcoins, etc. I dropped out from the scene fairly early since I was a True Believer and became quite disenchanted with the whole thing. All I was there for was "digital p2p cash" - I never once foresaw it as a major store of value.
Feels less like "ai cracked crypto" and more like having an insanely patient technical friend sitting next to you for 12 hours doing digital archaeology.
One can't help but observe the contrast between counterparty behavior here vs. the crypto boom. A counterparty service had a secret worth 400K in-hand and just passed it back to the user. Meanwhile crypto worries about MEV, the Dark Forest, etc.
This cycle is hostile in lots of ways, but the trustworthiness and absence of hostility in this dimension is quite nice.
Claude found an old wallet and then ran btcrecover on that. The question is why the user could not find an old wallet with any numbers of Unix tools himself.
Since we are dealing with Anthropic, the entire story could be staged of course.
Claude found a file on the computer that the wallet owner had not found. Claude didn't crack a password or do anything magic, it just searched for a file that the wallet owner had not thought to search for before.
So, where the wallet owner had previously only tried to access /Users/example/wallet.dat, Claude thought, "why don't I check if there is another wallet.dat file elsewhere on the system?" which it did.
The outcome is the same, it is great that Claude tried something that the wallet owner hadn't tried, but this is more an example of how dumb humans can be rather than how smart Claude is.
The trillions of passwords are a red herring and unrelated to the solve.
Yesterday I had ChatGPT walk me through fixing my single-node k3s cluster. It required rebuilding the sqlite database (while skipping a few corrupted records), then clearing the containerd cache, and then finally deleting a somehow-corrupted Secret record, and then recreating it.
Without it I would have given up way earlier, but the infinite patience to keep slurping in error messages and continue to troubleshoot really worked out.
Yeah Claude is really, really good there. You tell it the distro and the problem and it will solve it. Saved me a lot of pain when it came to swapping out an encrypted boot drive and was good about emphasizing the order of operations required for what I would consider a higher risk/complexity situation.
On a purely technical level - cool. But I still cannot get over the impression that even in this case LLMs show us how they are mainly useful to grifters. I mean, 5 Bitcoins worth 400K USD. Why? What intrinsic value does Bitcoin deliver? It's like trading for monopoly money.
> bot tried 3.5 trillion passwords before decrypting an old wallet backup
> After finding a mnemonic that actually turned out to be their old password a few weeks ago, the user dumped their entire college computer files in Claude in a last-gasp effort. The bot uncovered an old backup wallet file that it successfully decrypted, while also uncovering a bug in the password configuration that was preventing recovery up to that point.
So it switched from brute-force searching passwords against a file, to brute-force searching files against a password?
"the user dumped their entire college computer files in Claude in a last-gasp effort."
Claude has limits that would make this simple statement be much more complicated-
Via Claude "So the chat upload file size limit is actually 500MB per file (not 30MB as many third-party sources claim - those appear to be outdated). The 20-file-per-chat cap and the 30MB-per-file limit in Projects remain consistent across plans.
The real constraint at any subscription level remains the context window - how many tokens Claude can hold in memory at once during a conversation. "
10 years ago a 500mb hard drive was not unheard of and he may not have maxed out his storage. Also, cloud storage is more prevalent now.
I must admit this does sound a bit sensational.
So you’re telling me he used a computer to search all the files and found the old backup? computer file systems are truly proving to be revolutionary technology!
“ Out of frustration, cprkrn then dumped their whole college computer into Claude. This was when the AI discovered an older backup file of the wallet from December 2019 hidden in cprkrn's data.”
… this dweeb had a file containing their seed in their backup, claude just searched through the files
Heh. I was going to let the context just float and see how many people tried to hide/kill it. Thus of course not realizing the actual irony in forcing away my comment that someone who didn’t know better might think was derogatory is exactly why a gay jailbreak exists.
I've tried Claude Code with another LLM, it's very good at doing tasks and figuring things out. So this made me wonder, even though we know how good Claude models is, maybe the true value is in the harness now?
I have a similar claude story (much less money though), with the IRS R&D tax credit. The auditing firm initially said we qualify for $0. But then I had claude analyze past R&D reports and our expenses and it found the problem. The auditor had miscategorized our company.
So claude drafted an email even pointing to the right Internal Revenue Code (IRS Law), and specify why we fall under a specific category. The auditor got back to me two days later admitting their mistake and said our company now qualifies for $8k in tax credits. And a few months ago, it identified items in our AWS that saved us $250 a month (paying for itself).
So now I joke that even if I have a claude max plan, I've still come out ahead financially.
> So now I joke that even if I have a claude max plan, I've still come out ahead financially.
This is no joke; for better or worse, I see a day when I'm paying a lot more for this and it will be a bargain.
By my estimation (guess) you won't actually need to spend that much because the models are already getting a point where they don't need to get a whole lot better to be extremely helpful across many domains.
And it looks like those very helpful capabilities will continue to transfer to smaller models as well, as architectures and training regimes continue to refine.
I can fairly easily imagine a world where the only people needing to spend a lot of money on models are those that are using them to solve truly novel problems. The rest of us will get plenty of use at reasonable costs for the typical day-to-day helpful stuff.
All we need is something like Qwen3-coder-next but at Kimi K2.6 ability so it runs on laptop workstation hardware and we are set...soon?
In 2023 GPT-4 was allegedly 1.8T parameters. In 2026 we have ~100x smaller models (10-20B) that handily outperform it, and can indeed run on a laptop.
It highly depends on the task. For math and coding, sure. But for knowledge tasks GPT-4 is wayy better than even SOTA ~100B models. For my knowledge test cases the lines get blurry at >400B
How does "outperform" translate to the propensity of an LLM to hallucinate?
There seems to be a mass delusion about how capable SOTA models actually are. That's my only explanation for how poorly I find them performing in basic knowledge tasks compared to how others describe their prowess.
I understand you to be implying that I shouldn't trust my perception that there's a meaningful difference in how much different models hallucinate. I will take that under advisement, but I am still interested in the answer to my original question.
>I understand you to be implying that I shouldn't trust my perception that there's a meaningful difference in how much different models hallucinate.
Nope. Also I'm not GP.
I am eagerly awaiting being able to run a strong local model. I'd hand Apple $5k right now for a Claude in a box. I know the cost might not be there now, just saying that is around my ideal price point.
$10k might even be worth it - but i'm assuming that the more expensive it is the beefier it is too, which also means more electricity.. and i already run ~6 computers/servers in my house. If a power surge happens i'm going to go live in the woods lol.
I would do the same but my issue is that the models are changing so fast, so I don't want to be left out of the next model cuz it only runs on an even newer GPU or something like that.
But maybe my limited understanding is thinking of this wrong.
I wouldn't worry about hardware.
I've run the latest local models over the last year, including the recent Qwen 3.6 30B A3B, on a 9yo GTX 1080 and 32G RAM I have lying around[0]. If I can do that I don't think hardware will be a problem for you in the near term. The only updates I've needed were to Llama.cpp when a new class of model was released.
[0]: In my case, I want to see how local models perform on limited hardware, sacrificing context size and intelligence compared to SOTA models, so I have to really limit my expectations.
> I would do the same but my issue is that the models are changing so fast, so I don't want to be left out of the next model cuz it only runs on an even newer GPU or something like that.
I think the same, and it's why i stopped caring about running llama/etc at home last year. That coupled with the models being dumb by comparison to SOTA really make me fine with waiting.
But in a year or two it's going to be difficult to resist at home, assuming the pace of improvement holds.
Focus on what’s actually required for your workflows.
Anything beyond that is just hobby, or continued education.
You can run 6-12 month old state of the art models for that type of money,
like, yesterday.
Yea, but i don't consider them good enough. I barely consider SOTA good enough.
I'm hoping that by the time the rugpull happens with SOTA (claude/etc) that at-home will be in the 4.7-5.5 range? We'll see.
They were good enough 6-12 months ago.
Maybe your tooling is what’s keeping you from your dream.
Uh... get a UPS?
I do, though they're not as bullet proof as you'd hope to my understanding. Hell i have one at the house level too - since i have an EV sitting behind that as well.
Are you in a region that doesn’t mandate grounded electrical systems?
(UPS is still a great idea for your expensive gear.)
In my anecdotal experience there is a huge gap between GPT-5-mini which hallucinates relentlessly and Claude Opus or the latest GPTs which are fairly reliable. I'm hoping that gap can be closed with improved approaches for small models and that good reliability is achievable for LLMs without requiring absolutely mammoth computing resources.
For what it's worth, I also used GPT-5.2 (via duck.ai) this year for questions about taxes and it was helpful — which makes sense because there's an abundance of material about taxes out there to be synthesized, so a text predictor trained in that domain should do well.
[sci-fi “AGI” scenario] What if those with elite model access philosophize in a way us mere mortals can’t understand, so the elites have to prechew the ideas for us to bring them to our level, and they control the narrative?
In reality now, curious about social implications generally. Does this go beyond problem solving? Maybe the intelligence per token you get via your free library card/membership is insufficient to compete with peers in dating/employment/etc. markets, thus puts you at disadvantage.
That isn't really philosophy, but rather doom and gloom theories. Control the narrative on what exactly, how I write a bootstrap script for my servers? Or what type of flower is in this photo. Not everything is politics luckily.
Real world AGI scenario:
that’s already how world financial markets and governance work,
and yes, the best of the best models
and $ for tons of compute
will, for now, remain at the top.
I've seen this day sometime in December and not only with Claude. Wish I was joking on some days, feeling exhilarated on others.
To be fair regarding taxes it should be that you get all exemptions by default and the other side telling/justifying you why you don't qualify, instead of you not getting the things you should because you're not sure how to interpret the law or don't know this or that rule exists. Taxes shouldn't be that difficult, and the US version of it seems to be behind even the one I have here in europe (where my taxes are done "for me" for anything non business owner related). If the government is calculating your taxes anyway, they should just give the number to you instead of asking you a number and you better have the same as us or you're guilty of something.
Feels like a system that is deliberatly made to be more punishing for those who can't afford the help or the education to figure things out.
I would agree with you on most situations (like 1040 personal income taxes especially).
But in this particular tax credit, there's no way for the gov to know automatically what percentage of payroll was spent in qualified R&D expenses, since it's day to day business operations. Which is why we are _forced_ to hire an outside firm and pay them thousands of dollars (when Claude did an even better job), just to analyze how much of our time qualified as R&D expenses.
The problem I have is that I am forced to have to find a firm to do this, and most firms won't even work with companies as small as ours. So then we're stuck and losing out on years of R&D tax credits at the moment, when I really don't need them anymore, to be honest.
Or punishing to those that don’t pay for software and services to the companies that lobby for it to be this way.
i saw a meme once like:
IRS> Pay your taxes!
me> ok how much?
IRS> idk you have to figure it out
me> ...ok
IRS> if you get it wrong you goto jail
I remember a different ending
me> so you don’t know how much I owe?
IRS> no, we do…
me> ...ok
IRS> if you get it wrong you goto jail
> it identified items in our AWS that saved us $250 a month
I gave Claude Code access to the Azure CLI and had it review usage. It found a couple servers that were underpowered, but it also found a lot of things that weren't being used at all. The total reduction was $40k a year.
Same... I had chatgpt go over my taxes (I do it myself) and it found a number of savings I qualified for
Can you explain the steps you followed? Did you just feed it the whole return?
Kind of. I first of all did the entire return with it. So we went step by step and yes I fed the forms one at a time. So I filled in 1040 as best I could. Then just asked it what to do next at each step. It helps I've done it before so most of the steps it returned were ones I've done before. However, it did mention several things that I had not heard of, and also some new taxes that I had to file due to some exceptional events last year. So all in all, a solid use case. This year I have an accountant, but it saved my butt this last year, and I will absolutely run through my accountants decisions with it. It has an encyclopedic knowledge and an immense capability to understand without getting tired.
What sorts of things did it find?
Well my tax situation changed a bit this year, so there were some new taxes I had to file. Additionally it helped with depreciation for real estate. Helped me identify the exact part on the appraisal where I would get the cost of the dwelling (vs the land), and then walked me through the various ways the IRS allows people to do this. Helped me do some tax planning for next year as well around mega backdoor roths. Not lifechanging, but useful.
very trustworthy of the system sharing your taxes with a third party
What's the threat model here? OpenAI gets my social security number and Sam Altman steals my identity? OpenAI leaves an S3 bucket open to the public and my filled out 1040.pdf gets leaked to the world?
Oh no, OpenAI knows how much money I make and they're going to send me ads! Ads that are relevant to my interests. How connivingly evil of them!
People's tax returns are essentially public (yes I know they're not allowed to disclose them). Didn't send the forms in with social security numbers.
This absurd concern for privacy is silly in my opinion. The moment something is submitted to the government it ought to be considered public. Even your social security number is essentially public for anyone who cares to find it.
I would not submit my bank account information to these services, or my passwords, obviously.
Honestly, tax returns should be public again. Would make everyone better behaved IMO. It was this way for most of American income tax history believe it or not.
To be clear, my information has already been part of several breaches anyway. What protects you ultimately is the law not information security. Of course this point is often lost on engineering / computer scientist types who don't understand how law works.
omg. this is assuming the gov is completely incompetent and leaky. it seems doge "mission accomplished" indeed.
I use AI to help me do my accounting (how to categorize and account for things). It pays for itself because I need to spend money for far fewer hours from an accounting firm each month to make sure that captialized expenses, depreciation, tax credits (we have Historical Tax Credits from restoring an 1880s building) etc, is put in the books properly. The AI gets it right a high enough percentage of the time that I only the to have a real accountant look at things once a quarter to make sure it's all OK. I used to have a dozen questions every month.
Claude Code is really good at stuff like this. The other day I tried to recover some images from an SD card that had gone bad. I used GetDataBack to recover files, but they appeared to be malformed and didn't open in image viewers.
I tasked Claude to analyze the files and figure out what's going on, and eventually we figured out that each file had a custom metadata header + thumbnail + actual image concatenated. I had it write a python script and was able to recover all the images with their metadata. It's nothing a human couldn't have figured out, but it was definitely WAY faster than doing it myself.
I've also used Claude in the past to figure out how to break into routers with locked down firmware. It's great at suggesting and trying different approaches.
I have a friend that just picked up a new consulting job resurrecting an ancient Windows desktop application. No source control, no tests. And it's spread out over a dozen different folders with names like "_old", "_new" and "dates". Claude's doing a tremendous job in getting him to grips with what is actually happening in the application, what's relevant, what's not, what's different. I think it's literally saving him days and days at work.
if your friend has access to the binary and can pull it out to different box, they might get a lot out of a ghidra mcp -> https://github.com/LaurieWired/GhidraMCP
I'm not well versed at reverse engineering binaries or interpreting C/assembly so ghidra MCP has been an absolute gamechanger for helping me write tools. Once my project is complete, I plan to learn how to do the analysis myself manually and have cc guide me along the way.
I think it would be interesting, once the dust has settled, to do a compare with a less expensive model (time, capital, compute) such as deepseek 4.
Any reason to expect that this wouldn't work 100%? It's not like the different LLMs providers are that technically different from one another.
no such thing as closed source software anymore, just fully open and not quite fully open nowadays.
> I have a friend that just picked up a new consulting job resurrecting an ancient Windows desktop application. No source control, no tests. And it's spread out over a dozen different folders with names like "_old", "_new" and "dates".
That doesn't sound very impressive. Not being tracked with a version control system is fixed instantly with a git init, git add ., git commit .no AI required.
Covering the app with tests is also something that requires no AI. At most, coding agents can generate characterization tests in broad sweeps, but we are talking about a delta between hand rolling and vibe-coding of a couple of days.
Where LLM shines is helping developers build up an understanding of what is in place. Running /explain on a codebase can quickly provide you with a high level summary of what's in place.
The relevancy here is that he's denied the git history, versioning, branches, implicit documentation that even bad source control practices would have given him.
That's what the comment is saying. In normal repositories, version control acts as a record of the momentum of the direction the product was taking. If it's just "_old" and "_new," the developer has to read and understand both, which I think is going to be far more time consuming than your estimation.
I'm sure data recovery companies are pretty pissed that slightly esoteric data recovery abilities are becoming more accessible for average software devs. They were charging an arm and a leg to remote in and run scripts.
They still have two important moats: (1) expensive hardware tools (even stuff like SATA write blockers are kind of expensive for what they are), spare hard drive collections to swap failed PCBs, etc and (2) the "nobody got fired for hiring us" edge similar to how everyone calls in Crowdstrike/Mandiant after an incident. If a suit-level manager finds out customer data was lost, they are going to want to call in an expert so they can immediately tell the customer they did, not have the same internal team try to figure it out.
As an aside to #1: The cool thing is in modern times the hardware tools have come down stupidly cheap in price. Even SD card recovery is (vaguely) in the right skilled hands in a pseudo-professional home lab these days.
https://blog.acelab.eu.com/pc-3000-flash-spider-board-adapte...
I did EXACTLY that last night. Was doing by hand for about an hour and got to a point where I didn’t feel competent anymore and asked Claude to take from where I was.
5 minutes later I had almost 3 hours of important footage recovered.
> Claude Code is really good at stuff like this.
A lot of "Claude Code is best at X" claims are probably user-selection bias.
The people saying it are often exclusively Claude Code users, not people who are actively benchmarking Claude Code against Gemini CLI, OpenAI Codex, GitHub Copilot, and other agent harnesses on the same tasks.
The claim may still be true for certain scenarios, but the evidence is usually anecdotal, not comparative.
When I hear "claude code one-shotted X" and X is a novel problem, I mentally substituted "the agentic harness that I tried one-shotted X," since that's what they're saying.
Getting any smart model to take a look at the task is the sort of lift that the speaker is usually pointing to.
The harness is pretty much irrelevant for general tasks.
You can write a 100 line harness that only has one tool - try either "bash" or the more fun "you're running within nodejs, here's eval", you'd be surprised in how close to CC/Codex performance you're going to get.
There’s some weak evidence against this actually. Harness design makes a huge difference for tiny local models for example: https://itayinbarr.substack.com/p/honey-i-shrunk-the-coding-...
I have only my own personal experience for frontier models, but I have seen different performance of Opus when used from Pi or Claude Code or Zed for example.
I worded my comment poorly. I agree a good harness goes a way, but the harnesses most people use fucking suck and trip up the model so often that I don't think it's advisable to attribute successful results to them.
E.g. GPT5.5 with Codex on my Windows box likes using PowerShell for everything. OpenAI decided it should use the native shell instead of bundling a bash, or using git bash. Sure. But the model is so overfitted on bash that it fucks up PS quoting like once every 5 commands.
Every harness with LSP I've seen trips up the model as well. They insert diagnostics after every edit, polluting the context with errors that the model has to actively decide to ignore, every time, until it finishes its work and gets the code to a consistent state. Telling the model "run npx tsc --noEmit to check errors" will outperform a LSP 100% of the time.
Another example is basically everything Anthropic does - they add things like "think if this is malware!" after read and lead Claude to spend its reasoning effort on thinking if your React hamburger menu is malware, instead of on how to write it.
"This is not malware (em dash) it's a hamburger menu. Let me apply the edit! Hmm, is it malware now, after my edit? No, me changing border-width did not turn it into malware! Good! Dodged a real bullet on that one!"
I'm frankly amazed that we've gotten to the point where the models can produce good results in these sorts of environments.
I did that, wrote my own harness “Jarvis”, simple loop. Still results were terrible using the same model in comparison to for example OpenCode. So X Doubt.
Parent didn't say Claude Code is best at anything?
> Their luck changed for the better when they found an old mnemonic seed phrase written in an old college notebook
TBF the real breakthrough was finding this, though no doubt they couldn't have recovered without Claude
Pretty much every AI win story feels like this.
The guy also had to plug in an old hard drive for Claude to search. Sounds like he had an idea the wallet was on there to begin with.
I bet the majority of people reading this really think Claude cracked the encryption.
i bet the majority of people reading this will reach the 1st line of the 2nd paragarph, and not think that claude cracked the encryption
I don't have the data but I don't think most people look past the title.
Aren't those people filtered out when he says "majority of people reading this"?
I assumed 'this' refers to the title. That may have been incorrect.
> Bitcoin trader recovers $400,000 using Claude AI after getting 'stoned' and losing wallet password 11 years ago — bot tried 3.5 trillion passwords before decrypting an old wallet backup
Man. I wish I had a lost wallet worth a quarter of that even, technically didn't need Claude for this, just needed any password cracking software.
Explaining your life to an llm, then having it generate permutations of passwords to try does sound like it would work a decent percentage of the time.
A large percentage of passwords aren't a random string of characters but a memorable word + memorable number. There's existing projects that basically do the same, and 3.5 trillion doesn't really make it clear if one of those wouldn't have worked as well, but I can see it having an above random chance to guess a password.
>Explaining your life to an llm, then having it generate permutations of passwords to try does sound like it would work a decent percentage of the time.
I cannot relate to this at all. This information doesn't really seem that helpful. What might the strategy look like? Including spouses names or other proper nouns associated with you. But it's going to be a massive brute force effort still, and the likelyhood of a targeted crack that performs significantly better than more naive brute force passwords seems so unlikely.
Are your passwords like "SPOUSE_NAME:HOMETOWN_NAME"? Even if so there are probably more people with dictionary words that can be brute forced faster. IT would have to be the case that more people use patterns like that compared to something a regular dictionary attack could crack.
The amount of times I've gotten told a password and it contains birth year or anniversary year, maybe child birth year, is insane. I'd say 9 times out of 10 it's that or a dictionary word.
The idea that someone (the NSA?) is training models on all of our collected info, and using that to predict all of our hidden information, is horrifying.
The best time to start using a password manager was 10 years ago. The second best time is now.
If any authority wants your data, a password isn't whats stopping them.
So the remotely running AI now can guess many of your past and possibly future passwords when somone else promps it to ? Seems handy!
I had a high school friend that died about 10 years ago from an over dose. He was always tech forward and had talked in the past about getting drugs from the dark web to sell locally.
I wasn't particularly close with him after high school, but he was an only child, and I can only imagine his (older) parents just tossed his computer. I wouldn't be surprised if he had had a few hundred BTC on there.
That’s assuming he was smart enough to move his coins out of Mt Gox or whichever, now hacked, exchange he used at the time. (I wasn’t!)
I have a lost wallet with about 300 Bitcoin sitting in a landfill somewhere. I tried out Bitcoin really early on and mined those over a few weeks. But they were worthless back then and I was burning electricity for "nothing" so I stopped. This was before that 10k Bitcoin pizza purchase happened. I have some regrets lol.
Don't worry, we were also giving out millions (or more!) of dogecoin as tips on Reddit. Can't really get hung up on crypto shit.
have you tried to get the city to excavate it ? lol
I'm really thankful I put my bitcoin in a time vault back in 2012 or so. It was inaccessible until about last year, and my $10 is now worth $100k.
Thank you MtGox.
Way back in the day when Bitcoin first came about, I once idly contemplated spending some time and money on it just because it was a very cool technology. At the time it was a bit of a hassle because you had to mine your own.
Then I was especially tempted years later after running into the MtGox booth at CES, and seeing how convenient it had become. I remember asking a guy at the booth if Satoshi was really still anonymous or if any insiders knew about him, and he said "no" but was bit surprised I knew about Satoshi. I guess Bitcoin was still quite niche then even amongst a technical crowd.
I considered buying a few bucks worth of bitcoin then for lulz, but I thought that money was better spent on beer lol.
I've never really regretted spending that money on beer rather than bitcoin, because I knew that even if I did, it would 100% have been on MtGox and I would have lost it in the hack anyway, which would have been even more bitterly frustrating.
A few of pints of beer >> years of regret.
> MtGox
Whew, that brings me back!
I still think about the Bitcoin my buddy paid me for his half of a pizza ~15 years ago... worth 6 figures now haha.
I have a cousin who received around 2 BTC after playing some cards with some people. Wasn't worth much at the time. He sold the coins immediately.
Better not to dwell on such things.
Yup. I was really close to buying $1000 (at $1) worth of BTC ages ago. I don't dwell because the stress of managing that through time would have eaten me away lol.
With that said, i do regret not at least mining/etc. Back then i could have mined in many ways, and getting into it as a hobby probably would have meant holding larger amounts of BTC in the long run.
I remember thinking about buying $100 (at $10). And then realizing I didn't actually know how to do it and didn't feel like looking it up or going through whatever steps to do that kind of transaction online, or worrying about getting scammed....
A friend lost £2000 worth of BTC in MtGox which is probably worth a fortune at today's prices. The last time I spoke with him he said there was some sort of lawsuit for victim compensation. How did you recover your funds?
They emailed us during the course of the lawsuit, I followed the instructions and they sent me maybe half my BTC in the end.
Nice, congrats. What's a time vault?
It's sarcasm.
Everyone who had coin in Mt.Gox lost it during a hack. A portion of that was returned to the users who had a loss about a year ago.
Yeah my 100 stolen bitcoins got me a cool $4k check from the settlement. Definitely made whole by that :|
I think you would have been much better off choosing the bitcoin option. You'd have gotten around 40 BTC back, I think.
I don't believe I had the option. I just filed my claim, and several years later a check showed up.
Perhaps you should log in and verify. Maybe you missed the step about setting up an account on an exchange to get it transferred.
There was quite a few steps...maybe you still have something coming.
Hm no, at some point you were asked whether you want USD or BTC, plus (later) whether you want it guaranteed now or whether you want to wait for more.
That's what I remember, anyway.
Likely in this case the time vault was the collapse of Mt Gox, which has now recently been paying back holders.
It's something that locks your stuff so you can't access it for a while.
Someone gifted me 87 bitcoins back when they were worth ~0. They are still in some wallet somewhere I guess, and I saved the password on a harddrive I threw out around the same time
For 9.7 million, I would spend a considerable amount of time trying to recover that.
> …recovers $400,000 using Claude AI after getting 'stoned' and losing wallet password 11 years ago
I had to laugh: the most Bitcoin story ever.
If he was stoned, he would have probably spent his three bitcoins on pizza anyway.
The first pizza anybody bought that way cost 10,000 bitcoin, over $billion.
>over $billion.
BTCUSD has been over 100k, but is not currently.
Oh, only a mere $810m today.
I was making a long edit in a crappy wiki UI and my browser froze. It would have taken a long time to redo, hours.
I didn't want to take the chance of force closing and losing everything. I used claude code to extract my text out of the browser internals and filesystem objects.
I did this a long time ago before the age of AI. Core dump and then run strings on it. Very low tech but very useful!
How do you use Claude Code to access your browser memory?
I'd have to go back and look. It was 100% vibe coded.
Here's what it did: https://pastebin.com/fFXRCJWs
The claude code session took several other failed attempts until I vibe coded it to this direction.
Guessing it gave JS to drop into console.
Many crypto wallets use a key derivation function (KDF) to add an amount of computation (and memory usage) per password tried - to mitigate brute force of weak passwords.
The increase in compute (decrease in brute-force cost) combined with price increases in many crypto tokens means brute-forcing old wallets can become worth it years after passwords were forgotten.
And of course even smaller, local AI models can now easily write optimized scripts to brute-force any given KDF function.
how can that possibly work while supporting offline backup & restore?
The compute power needed use to be of the order of 5s per password try. So it effectively mitigate brute force back them, you need a absurd compute power to crack them.
Moore law did its thing, now you can do it with a lot less computer power.
Still, they didn’t give up because that wallet contained 5 BTC; this may not sound much, but it has a value of almost $400,000.
They are really underestimating their audiance here.
They are not though, because this audience is apparently eager to gobble up fake clickbait.
I'm no expert but using an old wallet with a changed password, and it working, seems like a major security design flaw.
In the physical world, I can't imagine too many people being happy that old keys to your house still work even after you've changed the locks.
Can someone more informed, help me understand how this worked and why it's ok.
I'm genuinely wanting to become more informed & better understand.
A wallet is just private keys of some specific public keys on the blockchain that have unspent output (UTXO). None of what’s described in this article involves the blockchain, only the storage and protection of the private keys on a local computer.
You can imagine that in your example, you didn’t change the locks on a house, but rather you put the house keys in a secure lock box and you changed the locks on this box.
Changing the locks on a house in this case means transferring from an old wallet to a new wallet and then abandoning the old wallet. That’s exactly what the OP is trying to do. It’s just that you need the original key to do it.
They didn't lose the key, they just didn't know which one is the correct one, where the lock is, and how the unlocking is done.
The wallet is akin to a lockbox holding your keys to the house. Breaking into the lockbox and changing it's lock does not affect the keys kept inside.
This sounds like an Ad, why would you tell the world about this? Too coincidential.
claude ads hot off the presses.
The amount of pure conspiratorial thinking will never cease to amaze me.
Lol great intersection of AI, Crypto and blind luck.
By getting stoned he was forced to hold until AI could solve his problem at a crypto high.
I spent a couple of days mining many years ago and got 2 bitcoins. At the time, they weren't worth the electricity they cost to mine and over time I lost the wallet and all information related to it.
I'd love to mine a bitcoin a day on my PC now...
My wife doesn't like it when I tell the story of the hard drive I threw away with a wallet with 2 dozen BTC on it.
But lets be honest - when BTC hit 100 bucks, we would have cashed it out thinking we were geniuses.
Yep, I timed the top of the market perfectly and cashed out 5 BTC at ~$80/BTC.
You are better than me. Back when you could mine BTC with the CPU, i had about 2 coins. I found it useless and silly and deleted my wallet at some point :)
Same here. There are a ton of unspendable coins for that reason.
For the record, they're still useless and silly, that just doesn't stop people from exchanging a nonsense amount of money for them.
>they weren't worth the electricity they cost to mine...I'd love to mine a bitcoin a day on my PC now...
you can!... but they wouldn't be worth the electricity now either. the cost of mining (amortization of hardware costs plus electricty) is the value of bitcoin. if bitcoins are a bargain to mine, more people will mine them thereby reducing rewards.
should you have mined more back then if you had magical perfect knowledge of the future? no: they weren't worth the electricity.
instead you should have bought more of them back then.
Think on the bright side, at least you didn't spend 10000 BTC to buy pizza... speaking of which, Bitcoin Pizza Day coming up in just over a week!
https://www.coinbase.com/learn/crypto-glossary/what-is-bitco...
There’s an interesting ethical question here.
The other day, I asked Claude to track down the leaked Claude Code source so I could study it. It refused, saying “given who made me, I’ll pass.” It gave me some pointers on how to find it myself, which worked.
There isn’t that much of a difference between “help me crack this bitcoin wallet” and “help me crack this executable.”
I don’t exactly have a solid point, just some general observations. First, I think we’ll see AI more and more simply refuse to do any kind of forensics, as forensics becomes more powerful. Second, that implies local models will become more valuable, since they’re the only ones willing to do that kind of work.
I once got myself banned from Claude by researching barbiturates, since they’re connected with suicide. So my third observation is that we’ll see an uptick in people getting punished for trying to do things with AI that people don’t usually do. (Luckily the unban form worked.)
Someone downthread asked “how’d he convince Claude the coins weren’t stolen?” Which is an interesting question, because presumably some people trying to crack a wallet have stolen it. So I guess the fourth observation is that the exact framing you approach an AI with will become more important. There was the classic “do this or I’ll cut off my arm,” which worked a year ago. But in the future it will be more like “hopefully the AI believes my story, or else I’ll get into trouble.”
It’s good there are multiple AI vendors, or else it’d get real dystopian real fast when the de facto AI’s policy becomes something you have no way of working around.
Kind of cool to hear... I had a couple computers running miners at home towards the end of my second marriage (which ended in 2010-2011), and after I had a few coins when realizing I couldn't actually spend them (IIRC $0.25 at the time), I just deleted the wallets. I had no foresight or faith that they'd be worth around $100k each in 15 years. I'm curious how many did the same, how many coins out there were just deleted altogether.
i distinctly remember a coworker many years ago offering to just give me like 10 bitcoin to see them and how it worked. i said "no, thanks". granted, hindsight is 20/20
I might have been your co-worker.
I went from giving them away to anyone in the office who asked (very very few), to selling them for under market when they were in the $10-50 range. Only three people took me up on the offer then, and they bought moderately life changing amounts I've later been thanked for. Such as a down payment for someone's first home after he got married. Luckily they held onto enough of their coins to make a meaningful difference in their lives like I preached about at the time.
The number of folks who ping me at various BTCH ATH hitting the news is expressing regret and laughing about it is... a lot!
It was strange being the "Weirdo Bitcoin guy" for a time. Then you started hearing about it in the mainstream news/etc. and it got even stranger to me.
I did quite well for myself, but of course wish I could go back in time and actually go balls to the wall on my early mining operation. I thought I was being irresponsible enough as it were - maxing out my rental townhouse's electric service with a rack of GPU miners in my basement, and some sketchy DIY electrical work to make it all happen.
Unfortunate to me cryptocurrency devolved into such a horrible place with the shitcoins, etc. I dropped out from the scene fairly early since I was a True Believer and became quite disenchanted with the whole thing. All I was there for was "digital p2p cash" - I never once foresaw it as a major store of value.
Feels less like "ai cracked crypto" and more like having an insanely patient technical friend sitting next to you for 12 hours doing digital archaeology.
Nowhere is it described as "AI cracked crypto".
Except your friend doesn't need to be paid and has limitless energy.
One can't help but observe the contrast between counterparty behavior here vs. the crypto boom. A counterparty service had a secret worth 400K in-hand and just passed it back to the user. Meanwhile crypto worries about MEV, the Dark Forest, etc.
This cycle is hostile in lots of ways, but the trustworthiness and absence of hostility in this dimension is quite nice.
Too many claude ads on this website
Claude found an old wallet and then ran btcrecover on that. The question is why the user could not find an old wallet with any numbers of Unix tools himself.
Since we are dealing with Anthropic, the entire story could be staged of course.
Anthropic has the private key now too
edit: Personally I don’t think they would take advantage of it, but still worth moving the BTC asap
The story is confusing some people.
Claude found a file on the computer that the wallet owner had not found. Claude didn't crack a password or do anything magic, it just searched for a file that the wallet owner had not thought to search for before.
So, where the wallet owner had previously only tried to access /Users/example/wallet.dat, Claude thought, "why don't I check if there is another wallet.dat file elsewhere on the system?" which it did.
The outcome is the same, it is great that Claude tried something that the wallet owner hadn't tried, but this is more an example of how dumb humans can be rather than how smart Claude is.
The trillions of passwords are a red herring and unrelated to the solve.
Does Claude turn out to be what 'Quantum' was promised; crack bitcoin? This could be fun.
Claude is also surprisingly good at analyzing system issues on a Linux system and solving them!
Yesterday I had ChatGPT walk me through fixing my single-node k3s cluster. It required rebuilding the sqlite database (while skipping a few corrupted records), then clearing the containerd cache, and then finally deleting a somehow-corrupted Secret record, and then recreating it.
Without it I would have given up way earlier, but the infinite patience to keep slurping in error messages and continue to troubleshoot really worked out.
Yeah Claude is really, really good there. You tell it the distro and the problem and it will solve it. Saved me a lot of pain when it came to swapping out an encrypted boot drive and was good about emphasizing the order of operations required for what I would consider a higher risk/complexity situation.
Not just Linux. Whiz at system management on Windows as well.
On a purely technical level - cool. But I still cannot get over the impression that even in this case LLMs show us how they are mainly useful to grifters. I mean, 5 Bitcoins worth 400K USD. Why? What intrinsic value does Bitcoin deliver? It's like trading for monopoly money.
> bot tried 3.5 trillion passwords before decrypting an old wallet backup
> After finding a mnemonic that actually turned out to be their old password a few weeks ago, the user dumped their entire college computer files in Claude in a last-gasp effort. The bot uncovered an old backup wallet file that it successfully decrypted, while also uncovering a bug in the password configuration that was preventing recovery up to that point.
So it switched from brute-force searching passwords against a file, to brute-force searching files against a password?
"the user dumped their entire college computer files in Claude in a last-gasp effort."
Claude has limits that would make this simple statement be much more complicated-
Via Claude "So the chat upload file size limit is actually 500MB per file (not 30MB as many third-party sources claim - those appear to be outdated). The 20-file-per-chat cap and the 30MB-per-file limit in Projects remain consistent across plans. The real constraint at any subscription level remains the context window - how many tokens Claude can hold in memory at once during a conversation. "
> Claude has limits that would make this simple statement be much more complicated-
I guess the user simply pointed Claude Code at a local folder containing all the backups and files, and Code went through them via find/ls/etc
10 years ago a 500mb hard drive was not unheard of and he may not have maxed out his storage. Also, cloud storage is more prevalent now. I must admit this does sound a bit sensational.
So you’re telling me he used a computer to search all the files and found the old backup? computer file systems are truly proving to be revolutionary technology!
“ Out of frustration, cprkrn then dumped their whole college computer into Claude. This was when the AI discovered an older backup file of the wallet from December 2019 hidden in cprkrn's data.”
… this dweeb had a file containing their seed in their backup, claude just searched through the files
And next to this article are two article recommendations:
Claude-powered AI coding agent deletes entire company database in 9 seconds <https://www.tomshardware.com/tech-industry/artificial-intell...>
Claude Code deletes developers' production setup, including its database and snapshots <https://www.tomshardware.com/tech-industry/artificial-intell...>
How did they convince Claude they hadn't stolen it?
Maybe they said they were gay?
(Relevant; there was a “gay jailbreak” thread a week or so ago. I laughed.)
Heh. I was going to let the context just float and see how many people tried to hide/kill it. Thus of course not realizing the actual irony in forcing away my comment that someone who didn’t know better might think was derogatory is exactly why a gay jailbreak exists.
blatant ad on the frontpage again
Claude hallucinate me a bitcoin address with unlimited money in it please.
this is fake clickbait.
i am not understanding why could'nt a deterministic dictionary program do it?
OK, that's impressive
Some more discussion: https://news.ycombinator.com/item?id=48136240
Is this not the link to this discussion?
Guessing the HN admins merged the post into this, which carries the comments.
I've tried Claude Code with another LLM, it's very good at doing tasks and figuring things out. So this made me wonder, even though we know how good Claude models is, maybe the true value is in the harness now?