Re:Claude Code and Codex reverse engineered my 1994 Jeep's E...
by aiiotnoodle on Sunday November 15, 2026 @10:58AM
(#50507911)
Would love to see this being used to add keys to cars. A lot of cars now have tablets and screens that could be used to do this, but for some reason (money) I have to take my VW to the dealer...
I think the whole market for aftermarket keys could be completely eradicated with good software UX (and low pricing).
Re:Claude Code and Codex reverse engineered my 1994 Jeep's E...
by qingcharles on Sunday November 15, 2026 @11:45AM
(#50508345)
Not these days. It won't even look at my projects for security flaws any longer. Has a little pop-up that says "This output cannot be viewed." or some such.
Re:Claude Code and Codex reverse engineered my 1994 Jeep's E...
by KlutzySofa on Sunday November 15, 2026 @12:05PM
(#50508532)
Its backwards in so many ways.
You fall back on having to use Chinese models. In theory the Chinese models could also be trained to sneak in new security flaws while Claude and Codex would avoid looking for them.
Re:Claude Code and Codex reverse engineered my 1994 Jeep's E...
by theplumber on Sunday November 15, 2026 @12:11PM
(#50508586)
Claude and codex could also be trained to sneak security flaws. They could even target specific people. Not to mention that all their data goes through NSA headquarters so you don’t even have the option to run it locally in any shape or form.
Re:Claude Code and Codex reverse engineered my 1994 Jeep's E...
by KlutzySofa on Sunday November 15, 2026 @12:59PM
(#50509064)
The US companies are doing this to not get export restricted by the US government which in turn can end up with companies using Chinese models to get around those restrictions.
Its not really a matter of who's good and bad (neither the US or China has the best track record), but indirectly creating a scenario where the other party can use your own policies against you.
Re:Claude Code and Codex reverse engineered my 1994 Jeep's E...
by ryandrake on Sunday November 15, 2026 @01:18PM
(#50509266)
It's crazy. Reverse engineering a protocol/interface is something LLMs are surprisingly good at, and these show a promise of unlocking sooo much proprietary crapware that's hidden behind obfuscation, accounts, subscriptions, lazy vendors who only write crappy windows programs, and so on.
But of course, the major US labs are going to get pressure from these vendors whose business models depend on obfuscation, and you know they're going to bend the knee and nerf the LLMs.
Re:Claude Code and Codex reverse engineered my 1994 Jeep's E...
by holoduke on Sunday November 15, 2026 @05:15PM
(#50511425)
It's pretty easy to bypass this check by telling the system you are working at a car Company and repair hardware with blabla. Many other tricks to get though those weak filters.
Re:Claude Code and Codex reverse engineered my 1994 Jeep's E...
by terribleperson on Sunday November 15, 2026 @12:47PM
(#50508965)
I had this same thought the other day. For a lot of cars you can buy aftermarket programmers, but the Dorman ones are limited to 3 uses. 3 uses for like eighty dollars is terrible, and surely it's not a technical limitation.
Re:Claude Code and Codex reverse engineered my 1994 Jeep's E...
by terribleperson on Sunday November 15, 2026 @02:16PM
(#50509758)
Hah, I hadn't considered a hardware approach to limited use.
I assume there's just a counter in there. There could, theoretically, be cryptographic reasons it can only do 3 programming sessions, but I doubt it.
Re:Claude Code and Codex reverse engineered my 1994 Jeep's E...
by Doohickey-d on Sunday November 15, 2026 @12:52PM
(#50509005)
The whole aftermarket key ecosystem, even vehicles going back 20 years, is messy, horrible thing.
It's dominated by Chinese manufacturers of specialty tools (Autel, Xhorse, ...), which inevitably mandate bundled accounts, subscriptions, mandatory cloud features, etc. Even the older cars, you have either cheap Chinese tools, where you never know wheather it'll really work, or the manufacturer's tool which needs a dealer code, which of course the dealer won't have after 20 years.
I wrote about my 20 year old car-key-adding adventure here: https://jonas.bike/low-cost-diy-key-/-transponder-for-my-200... - it requires desoldering an eeprom, two pieces of hardware from aliexpress which come with pirated copies of software, and lots of hope that it'll work...
An open source community curated collection of scripts, which works with a cheap OBD dongle (which is all the chinese devices are: OBD hardware + per-car software), would be amazing.
Re:Claude Code and Codex reverse engineered my 1994 Jeep's E...
by londons_explore on Sunday November 15, 2026 @02:07PM
(#50509677)
Frustratingly those cheap OBD dongles (ELM327 clones) seem to be mostly designed for diagnostics. They effectively run a regex over all the canbus data and tell the computer what matches.
However they can't match and forward everything (not enough RAM) and miss loads of packets, which makes them useless for things like reflashing firmware or interfacing with OEM tooling.
Re:Claude Code and Codex reverse engineered my 1994 Jeep's E...
by holoduke on Sunday November 15, 2026 @05:11PM
(#50511390)
So what odb should I use to start reverse engineer my TCU from a old 2006 ferrari? These tcus are hard to come by. I wish I could replicate it's behavior. Not sure if that is possible. I think i can do it
Re:Claude Code and Codex reverse engineered my 1994 Jeep's E...
by aiiotnoodle on Sunday November 15, 2026 @06:36PM
(#50512069)
Great read and similar to the same route I went down. I also got the xhorse for a corsa, was able to program the transponder but wasn't able to get the lock/unlock central locking buttons working using xhorse. I ended up just living with it and sold the car with it as a "spare" key.
The corsa also needed a "security code" that wasn't with the log book like it was supposed to be and would have required a trip to the dealer to get butI read about ripping the eeprom off the back of the dashboard to get the code but decided to cut my losses.
Re:Claude Code and Codex reverse engineered my 1994 Jeep's E...
by bri3d on Sunday November 15, 2026 @01:31PM
(#50509375)
Keys in European and newer American cars reliably have Some Kind of real-ish crypto scheme applied, as this is fairly regulated in Europe. It’s an anti theft measure driven by the real stolen auto trafficking issue faced especially in Europe. Most of the aftermarket tools vendors use some kind of exploit (code exec, arbitrary read, update signature bypass, etc.) to extract the crypto material from a control module participating in the immobilizer - it’s neither a weird conspiracy nor as simple as it seems to bypass.
Re:Claude Code and Codex reverse engineered my 1994 Jeep's E...
by ronnier on Sunday November 15, 2026 @02:32PM
(#50509896)
You can do that with teslas. Buy two keys (they at those credit cards sized keys) and you pair them on your own. Two for $40. Do it right at home. You can pair multiple cars to a single key. Or make guest keys
Re:Claude Code and Codex reverse engineered my 1994 Jeep's E...
by ronnier on Sunday November 15, 2026 @03:19PM
(#50510289)
But you have to buy an expensive key. I have two Teslas and a Toyota and my Toyota is so annoying that I even have to have a key. My Tesla just uses my phone (the credit card keys are totally optional)
Re:Claude Code and Codex reverse engineered my 1994 Jeep's E...
by misterderpie on Sunday November 15, 2026 @11:05AM
(#50507980)
LLMs end up changing the consumer market not the way big corporations would like to. It is impressive to see that we'll be able to give ownership back to the customer, or enable others to do so.
Very nice demonstration.
Re:Claude Code and Codex reverse engineered my 1994 Jeep's E...
by close04 on Sunday November 15, 2026 @11:16AM
(#50508078)
Wait until those big corporations start using LLMs to learn how to lock things down even further, or implement checks that disable your car when it detects alterations to the software “for safety reasons”, or go after drivers who do these things for loss of revenue, etc. A ‘94 Jeep will be fair game but a new car might not be.
Re:Claude Code and Codex reverse engineered my 1994 Jeep's E...
by SV_BubbleTime on Sunday November 15, 2026 @11:22AM
(#50508136)
New cars already have it.
Lockstep processors, hardware security modules with immutable unreadable keys, and PKI keys that the mfg holds the private for that can only be accessed on their online services.
It’s locked down now and this started in 2018 or so. It’s just worse now.
To automotive, there is really still no right to repair. They get around it by selling “3 day passes” to their service, but with their hardware and at their prices.
They’re all moving the Tesla system of more centralized brain systems which will be even more protected.
Re:Claude Code and Codex reverse engineered my 1994 Jeep's E...
by misterderpie on Sunday November 15, 2026 @11:28AM
(#50508173)
Sad to hear. I am not a car owner, but lived in the EU for most of my life. The consumer market owes quite some consumer rights to the EU, so I guess all we can do is hope that they'll eventually crack down on these measurements again.
Re:Claude Code and Codex reverse engineered my 1994 Jeep's E...
by abletonlive on Sunday November 15, 2026 @12:56PM
(#50509051)
> The consumer market owes quite some consumer rights to the EU
Let's be clear that it's just incentives. If the EU economy was based on creating companies and new products there would not be as many incentives for consumer "rights" there.
Re:Claude Code and Codex reverse engineered my 1994 Jeep's E...
by PorciiVorbesc on Sunday November 15, 2026 @02:01PM
(#50509627)
> The consumer market owes quite some consumer rights to the EU
Not for auto owners. Auto makers and dealers get away with a lot of shady shit fucking over their customers in the EU all the time.
Like sibling said, there are incentives to regulate stuff that has no impact on EU businesses, shit like USB-C or replaceable batteries, but they won't touch the auto industry to mandate easier repairs and less lock-downs.
Re:Claude Code and Codex reverse engineered my 1994 Jeep's E...
by sam1714 on Sunday November 15, 2026 @05:47PM
(#50511702)
I'd also add the big driver for repairability is the US EPA which under Clean Air Act powers requires any emissions-related repair to be able to be done by third parties. They have mandated SAE J2534 for module updates and publicly available service manuals, for example.
Pure battery EVs are considered to have no emissions-related components, which is why Tesla gets away with that they do.
Re:Claude Code and Codex reverse engineered my 1994 Jeep's E...
by SV_BubbleTime on Sunday November 15, 2026 @09:54PM
(#50513465)
J2534 is just a hardware interface.
You want to use it? Go ahead and buy a 1,3,7,30 day pass to the mfg’s service. Then, good luck with your J2534 because they only approve a couple, and better find a windows machine because their .NET program only runs on Windows 10.
Re:Claude Code and Codex reverse engineered my 1994 Jeep's E...
by knowaveragejoe on Sunday November 15, 2026 @03:51PM
(#50510606)
I'm hoping the slate truck will be different, and that whatever problems emerge out of giving consumers real control over the hardware they purchased doesn't hamstring the company somehow.
Re:Claude Code and Codex reverse engineered my 1994 Jeep's E...
by close04 on Sunday November 15, 2026 @11:36AM
(#50508258)
Yes, they have been on that road for years. Tesla showed everyone the way with things like not being able to change a headlight with another original one without Tesla’s blessing. What I meant is that whatever I can do with an LLM they can do more. And if they can invoke “safety” that'll have the law on their side for some time.
Re:Claude Code and Codex reverse engineered my 1994 Jeep's E...
by close04 on Monday November 16, 2026 @04:05AM
(#50515543)
> we have the benefit of unlimited time, having physical control, the analog hole.
We do, but everything is "software defined car" now and loopholes usually get closed with mandatory software updates done with the mandatory inspections. They put everything in one basket together with safety critical systems because that justifies any lockdown.
> Right to repair laws laws usually give you the right to use non-OEM parts.
Manufacturers already make that very expensive. Mechanics have to be certified, they need special and very expensive tools, everything is wrapped in cryptography and the manufacturer has to be in the middle of everything, for a price. [0] So go ahead, use your OEM part, you'll still pay thousands for the mechanic that has the necessary prerequisites and for the manufacturer's services.
I pray I'm wrong but we need strong regulation and enforcement, and the manufacturers can still maliciously comply.
Re:Claude Code and Codex reverse engineered my 1994 Jeep's E...
by close04 on Monday November 16, 2026 @04:30AM
(#50515685)
Nothing prevents you today and technically. What prevents most owners is that they lose a lot of big features they paid for. Who wants to pay for the full car and rip out much of it? This isn't an acceptable solution for most people. They might as well not buy the car at all. And even the technical possibility can change so easily in the future.
Re:Claude Code and Codex reverse engineered my 1994 Jeep's E...
by CamperBob2 on Sunday November 15, 2026 @12:07PM
(#50508550)
They are now required to do just that, if they want to sell to the EU market. When you hear the term "cybersecurity," the real subject is the protection of revenue streams from attack by end users.
Re:Claude Code and Codex reverse engineered my 1994 Jeep's E...
by damowangcy on Sunday November 15, 2026 @12:15PM
(#50508651)
>big corporations start using LLMs to learn how to lock things down even further
Why would they? They just need to fund lobbyist to make it illegal for any AI to reverse engineer anything copyrighted or with IP.
Sure, you can use open source AI to do it but not everyone will/can do it.
So, they don't need to ban everyone, they just need to make sure the final code isn't available freely on GitHub. That would minimize the damage, it's not a perfect solution but more than enough to protect the company.
Re:Claude Code and Codex reverse engineered my 1994 Jeep's E...
by damowangcy on Sunday November 15, 2026 @11:04PM
(#50513880)
Yes, but the more sophisticated it is technically, the more expensive to create and maintain while only making it harder, not impossible, to break.
Unless your company's moat is in the thing that was reversed engineer, I don't think most company cared enough to spend huge amount of resource to prevent it technically. The only thing that they need to prevent is the liability that comes after the product is used inappropriately or illegally.
And it you can defend something with AI, there's no reason why you can't attack it AI.
Re:Claude Code and Codex reverse engineered my 1994 Jeep's E...
by baby_souffle on Sunday November 15, 2026 @01:02PM
(#50509093)
> We have maybe 5 years of freedom before the walls close in around us, so take advantage of the Wild West while you can.
I suspect you're right BUT still maintain a bit of optimism.
We miiiight end up in a future where most hardware comes with an OK-ish app/interface and some sort of LLM-Friendly technical documentation so if you don't like $aspect of the device, you're free to roll your own after indemnifying the manufacturer via something like an eFuse.
Re:Claude Code and Codex reverse engineered my 1994 Jeep's E...
by testtestmetest on Sunday November 15, 2026 @01:31PM
(#50509377)
Lmao if a company starts shipping llm documentation for its products so you can tamper with them, you should assume youve woken up in a parallel universe.
I think most people on here fail to realize the demand to customize and build things hasnt gone up. Just because llms exist now doesnt mean normies all of a sudden want to start getting into programming, they dont care.
The only number thats gone up is people who were slightly interested in programming/CS before but were too undisciplined to learn things themselves. Llms are great for those lazy people.
Re:Claude Code and Codex reverse engineered my 1994 Jeep's E...
by Crye on Sunday November 15, 2026 @02:33PM
(#50509905)
I get the point that you're trying to make, but it's not just people who were undisciplined. I spent many years developing code projects that I remember spending hours every afternoon just to ship a single feature, and that pace has obviously changed.
I still get the joy of translating ideas and orchestrating and architecting solutions to problems I have in real life, but just at a quicker pace.
If some of those tools and solutions can help other people, that's great. If not, that's okay too. I hope everyone sort of has that mindset who's building today.
Re:Claude Code and Codex reverse engineered my 1994 Jeep's E...
by Caloric3002Swed on Sunday November 15, 2026 @03:37PM
(#50510447)
>The only number thats gone up is people who were slightly interested in programming/CS before but were too undisciplined to learn things themselves. Llms are great for those lazy people.
Re:Claude Code and Codex reverse engineered my 1994 Jeep's E...
by leesalminen on Sunday November 15, 2026 @03:55PM
(#50510637)
My dad, who is in his 70s, recently started recreating a piece of audio visualization software from the 00s using Codex. He’s having the time of his life! He was never interested in programming/CS. He just wants his software back.
Re:Claude Code and Codex reverse engineered my 1994 Jeep's E...
by randomaccountpe on Monday November 16, 2026 @12:38AM
(#50514412)
I mean, sort of agree/disagree with the sentiment.
I spent years and years learning coding, but when I finally got old enough, I realized I needed to graduate uni, get a job, get money, and get out of my parents house ASAP.
So I had to hone all the things I'd learned and just focus on one path. (Over generalization but whatever front/back/fullstack is, as that's what a majority of the jobs are).
I was always going to have discipline for things related to my job, but no way in hell was I ever going to learn: The insides of my car, am I getting scammed on the price of tyres, what's the best diet/nutrition, how to fix a door, how does the AC work.
Even on the job, what's the best way to do XYZ. Well, I want the best way possible, but that's just not going to happen. (Talking about pre LLM coding).
With all the new LLMs and coding agents, I'm quite happy to say, that I can figure out how to do all these things much more lazily, and efficiently. I definitely see a future where I can now learn to fix my own car, my door, etc.
Re:Claude Code and Codex reverse engineered my 1994 Jeep's E...
by testtestmetest on Sunday November 15, 2026 @01:34PM
(#50509398)
No they wwont, they'll never be maintained. Theyll decay with their vibes. People who vibe code whole swaths of software are'nt capable of caring enough to maintain it. Llms mostly make smart people lazy and or stupid.
Re:Claude Code and Codex reverse engineered my 1994 Jeep's E...
by colonelspace on Sunday November 15, 2026 @03:21PM
(#50510306)
I used Claude to document the wire and firmware protocol for an expensive piece of equipment I own where the only software available ran on PowerPC OS X.
Is this just "vibes"? Maybe, but I now have a full, verified spec, against which to build software for myself, also with a little bit of "vibes".
Re:Claude Code and Codex reverse engineered my 1994 Jeep's E...
by jdiff on Sunday November 15, 2026 @08:08PM
(#50512790)
How much is vibe verification worth?
I'm currently teasing apart numerous ancient Text to Speech synthesizers with Sol and the amount of times I ask a follow up question that forces it to correct 5 things it's already said/documented is ridiculous.
Re:Claude Code and Codex reverse engineered my 1994 Jeep's E...
by freecodeio on Monday November 16, 2026 @05:39AM
(#50516113)
I'd say 100%, since it's a spec? what's a spec worth if we don't know how accurate it is? this isn't exactly a post apocalyptic world where even junk is valuable
Re:Claude Code and Codex reverse engineered my 1994 Jeep's E...
by jdiff on Monday November 16, 2026 @07:15AM
(#50516775)
You would have tell you. My point is that it's clearly less than a true verified spec. Even in a harmless area like mine, it's at least horribly obnoxious. Not worth too much to me because at the end of the day I do have papers and open codebases to study.
But this is your unspecified expensive device. A verified spec wouldn't have you send a command or configure a table that bricks or damages it. A vibe verified spec very well might. How much is that worth to you? What are you willing to lose?
Re:Claude Code and Codex reverse engineered my 1994 Jeep's E...
by testtestmetest on Sunday November 15, 2026 @06:23PM
(#50511972)
I wouldnt say implementing an official spec with an llm is vibe coding.
Hard to judge though, most people are just winging it though and shoving it out the door as if its impressive.
But yes, I think its probably bad because you probably didnt learn all that much and became more reliant on Anthropic. Unless you dont care about learning, then who cares.
Re:Claude Code and Codex reverse engineered my 1994 Jeep's E...
by holoduke on Sunday November 15, 2026 @05:04PM
(#50511333)
Vibe coded software is currently already better than half the developers out there. Accept it. Code won't be seen anymore. It's a skill only a handful of jobs require to have.
Re:Claude Code and Codex reverse engineered my 1994 Jeep's E...
by testtestmetest on Sunday November 15, 2026 @08:17PM
(#50512847)
Actually, most things dont run in a browser. I know you probably live in that world but its not a handful.
Just because you've chosen to turn off your brain and prompt your way to whatever, and your 10 MAU app works fine does not mean large systems, critical devices and critical infrastructure's code will no longer need reviewed. Sorry your dad's pacemakers firmware code, and the ancillary systems that enable its development will need its code reviewed. Along with aircraft, satellites, municiple traffic systems, countless medical devices ect ect. I could go on forever.
Not everything is non serious SaaS platforms with little users. Just because you have success writing slop that exists under no load, and nothing relies on it doesn't mean the world is now different.
You have SEVERE dunning kruger syndrome, and unfortunately a lot of other people have this same ailment and the worlds systems are decaying as everyone turns off their brain and offshores everything difficult to language models, because humans like you are lazy and will always try to get the free lunch that doesn't exist. People will die because of this attitude.
Re:Claude Code and Codex reverse engineered my 1994 Jeep's E...
by cortesoft on Sunday November 15, 2026 @03:57PM
(#50510682)
We still have amazing freedom on the web like we never did before. They have been trying to stop online piracy for 30 years and have no succeeded at all.
Re:Claude Code and Codex reverse engineered my 1994 Jeep's E...
by InsidiousWombat on Sunday November 15, 2026 @08:35PM
(#50512974)
I don't see the genie (or maybe pandora's box) going back in once people started being able to run reasonably powerful models at home (if at pretty high cost in hardware).
stopping it would require outright putting DRM on the GPU and any other hardware that can run it, good luck with that (and not having the LLM itself be used to break out of it)
Re:Claude Code and Codex reverse engineered my 1994 Jeep's E...
by stephbook on Sunday November 15, 2026 @03:16PM
(#50510266)
You extrapolate from current machines. I do believe that AI-assisted companies will build AI-assisted always-online machines that will be worth their money, no need to hack.
I imagine a $25 Starlink subscription for a John Deere fleet of autonomous tractors that work through the night will replace what we have now.
(to stay within physical machinery, like the post.)
Re:Claude Code and Codex reverse engineered my 1994 Jeep's E...
by agumonkey on Sunday November 15, 2026 @03:30PM
(#50510382)
some 'funniest ways to use claude ..' youtube videos felt that way.
people were able to revive, repair, optimize a lot of day to day devices, very nice view. that said cars and heavy appliances.. should be done in a big public warehouse safety lab
Re:Claude Code and Codex reverse engineered my 1994 Jeep's E...
by muststopmyths on Sunday November 15, 2026 @11:14AM
(#50508055)
can someone explain the root cause of the smog test failure to me ? it sounds like the car was using a California smog test-specific fuel map at the speeds of the test. Wouldn't that make it pass the test, not fail it ?
Re:Claude Code and Codex reverse engineered my 1994 Jeep's E...
by threetonesun on Sunday November 15, 2026 @11:34AM
(#50508241)
Yeah doesn’t really make sense, there’s very likely a mechanical issue the LLM “fixed” with a richer fuel mix. But also it’s a 94 XJ the reason the CA map was there was probably because they got abhorrent gas mileage without it.
Re:Claude Code and Codex reverse engineered my 1994 Jeep's E...
by bflesch on Sunday November 15, 2026 @11:57AM
(#50508456)
You can adjust the mixing ratio of fuel and air in a way so that the exhaust has reduced emissions. This will reduce performance of the car but allows the exhaust to pass the emission tests.
Re:Claude Code and Codex reverse engineered my 1994 Jeep's E...
by eightysixfour on Sunday November 15, 2026 @07:24PM
(#50512463)
His particular car had a tune to reduce emissions by running lean at certain RPMs. While that reduces some emissions, it increases NOx, which is what was failing. You can add some more fuel to get closer to lambda, reduce the NOx, and your emissions will have more CO and HC.
Re:Claude Code and Codex reverse engineered my 1994 Jeep's E...
by qingcharles on Sunday November 15, 2026 @11:56AM
(#50508451)
Definitely one of the Claude video skill stacks that are going around. Dude should make a whole separate video about the animations, he'd get as many views again just for that.
Re:Claude Code and Codex reverse engineered my 1994 Jeep's E...
by dghlsakjg on Sunday November 15, 2026 @11:33AM
(#50508232)
It also didn’t solve the root cause of the issue. ECUs don’t need to be overwritten to get stock performance and emissions. Something is wrong with hardware/mechanical that needs to be solved
Re:Claude Code and Codex reverse engineered my 1994 Jeep's E...
by d1l on Sunday November 15, 2026 @01:38PM
(#50509428)
Yeah I had the same impression. There was plenty of data presented and code was dumped and reversed…but was the actual issue solved? What even was causing the issue? Presumably it passed smog sometime in the past couple years so what degraded? It was lean and there was some kind of issue with pressure. I can’t imagine a reflash or whatever is the fix.
It definitely feels like a classic case of the operator reinforcing Claude to keep pursuing a goal that’s counterproductive to a robust solution. Claude complied and they went hard on it but I can’t imagine the driver is any better off and more likely in a worse spot as the underlying issue wasn’t solved.
Re:Claude Code and Codex reverse engineered my 1994 Jeep's E...
by dghlsakjg on Sunday November 15, 2026 @09:35PM
(#50513331)
To be fair to AI, I put his symptoms into Claude and it suggested all of the obvious things and steps to start with. None of them were to remap the fuel delivery system.
Re:Claude Code and Codex reverse engineered my 1994 Jeep's E...
by mrbombastic on Sunday November 15, 2026 @06:04PM
(#50511837)
”You attach an assumption to a piece of evidence, you start to bend the narrative to support it. Prejudice yourself.”. I find by default while claude is great at analyzing logs and diagnostics from a variety of places it is really prone to this bad detective pattern of choosing an explanation and sticking with it even when evidence starts to bend away from it. you really have to play the role of the skeptic to do do good root cause analysis. Which is very hard to do in unfamiliar domains.
Re:Claude Code and Codex reverse engineered my 1994 Jeep's E...
by Rebelgecko on Sunday November 15, 2026 @12:13PM
(#50508625)
LLMs have been pretty good at breaking bad crypto in cars. In my Hyundai, they were able to figure out the right incantation to get the car to run unsigned code off a USB stick. Although I guess we're in trouble once carmakers use LLMs to lock things down even more
Re:Claude Code and Codex reverse engineered my 1994 Jeep's E...
by intrasight on Sunday November 15, 2026 @12:38PM
(#50508876)
Most shop security is security through obscurity. For them to lock it down, they need to manage those secrets all the way back to a cloud identity provider. They can and probably will do that eventually.
Re:Claude Code and Codex reverse engineered my 1994 Jeep's E...
by bri3d on Sunday November 15, 2026 @01:26PM
(#50509331)
Old news - for 20+ years in Europe, 5-10 years in the US, and now in Korean and Japanese cars too, ECM updates have been cryptographically signed for warranty, emissions, and now EU cybersecurity regulatory compliance. It turns out there are still bugs in things with cryptography too though.
> Claude suggested that it would run the fuel pump on its own. Um and I said, "Hey, if you can do that, why not? I don't think you can do that." And it started priming the fuel rail and it kept priming it and kept priming it. The fuel pump was on for about 10 seconds and I started to freak out uh and requested it to stop because that was the first time that Claude had actually done something inside of the computer to have a change.
Although we've seen this play out with handing cloud keys to Claude and it just going off on its own, seeing it poke around and make physical, real-world changes in its clumsy, cavalier way is both awesome and terrifying.
Re:Claude Code and Codex reverse engineered my 1994 Jeep's E...
by terribleperson on Sunday November 15, 2026 @12:45PM
(#50508951)
I'm kind of surprised Claude decided to control the fuel pump directly. I was having Sol help me get control of a CNC mill, and it was pretty adamant about always building tools for every attempt at control.
Re:Claude Code and Codex reverse engineered my 1994 Jeep's E...
by BikiniPrince on Sunday November 15, 2026 @01:10PM
(#50509180)
It usually builds throw away tooling. It’s not polished or user friendly. If you give it a directive to preserve and document tooling it will reuse it later and potentially improve it.
Re:Claude Code and Codex reverse engineered my 1994 Jeep's E...
by terribleperson on Sunday November 15, 2026 @02:14PM
(#50509743)
I was more surprised by the apparent lack of guards. Sol recognized that controlling a CNC mill has both human-safety issues and risks to the machine, and explicitly added various guards (like having the program stop movement after a certain amount of time, distance traveled, or lack of communication). It also initially started out with like 1mm moves and I had to instruct it to do longer moves so I could actually observe it. I didn't even have to mention safety once. Claude deciding to turn on the fuel pump without a test run where it just turns it on for one second is surprising. I'm also surprised the user went along with that, but content I guess?
Re:Claude Code and Codex reverse engineered my 1994 Jeep's E...
by georgemcbay on Sunday November 15, 2026 @03:58PM
(#50510686)
I've had Claude be very cautious about doing far less dangerous things than running a fuel pump.
This isn't meant as a defense of Claude or to refute the OP anecdote, but rather to point out that all models are non-deterministic even when safeguards are meant to exist.
Just because they don't do the stupid thing 90% of the time doesn't mean they won't do the stupid thing 10% of the time.
Re:Claude Code and Codex reverse engineered my 1994 Jeep's E...
by 01HNNWZ0MV43FF on Sunday November 15, 2026 @02:20PM
(#50509784)
Mine's never said such a thing, that's a good idea though.
I notice with mine that it only matches my level, by default it's a dump slop generator, then if I start asking, "shouldn't you test this, shouldn't you refactor that" suddenly it 'remembers' all kinds of software engineering knowledge.
Re:Claude Code and Codex reverse engineered my 1994 Jeep's E...
by neya on Sunday November 15, 2026 @01:02PM
(#50509094)
We are used to Claude refusing to reverse engineer a lot of stuff. But, we are not used to Claude refusing to stop while reverse engineering something. That would be truly terrifying. You can't win - you cut the connection, you risk bricking the device, you let it continue, you're looking at Terminator, perhaps.
Re:Claude Code and Codex reverse engineered my 1994 Jeep's E...
by rpcope1 on Sunday November 15, 2026 @02:11PM
(#50509718)
The guy kind of seems like he might be in a little over his head. That said I have used Claude (and other LLMs) for some similar stuff as an experiment and have seen them mess things up. Claude jacked up the canbus on a pre can gateway car while it was rolling down the road, which was funny because the whole Christmas tree in the dash lit up and my gauges dropped out (note: brakes and steering are mechanical and I knew exactly what was happening). I've never seen it so insistent that it needed to phone home and cry to Dario when it figured out it managed to crash a bunch of shit like the ABS and TCS, even though a restart cleared it and I told it that this was on a controlled safe course.
In general I've seen them split between 30% brilliance, and 70% pants of head outrageously stupid shit for electronics and automotive stuff (both diagnostics and repair/modification). I would not suggest even frontier models if someone didn't understand what they were doing.
Re:Claude Code and Codex reverse engineered my 1994 Jeep's E...
by pcthrowaway on Sunday November 15, 2026 @04:10PM
(#50510793)
> In general I've seen them split between 30% brilliance, and 70% pants of head outrageously stupid shit for electronics and automotive stuff
I assume the corpus of text they were trained on matches that 30/70 split?
I assume there's some selection or fine-tuning that happens to improve models fed similarly unreliable training data for other domains... curious how this works.
Re:Claude Code and Codex reverse engineered my 1994 Jeep's E...
by rpcope1 on Sunday November 15, 2026 @08:59PM
(#50513125)
I mean, I've seen some stupid it's told me at various stages that would never show up in a textbook like Halderman, and would get you ridiculed on a forum, and more importantly were sufficiently stupid that even a lube tech wouldn't do it. I like trying to push and see what the technology can do or suggest, but it will suggest outrageous shit. As an example I had it tell me that I should put something like ten quarts of oil in a gas engine (don't recall what). It was probably the right amount for a Duramax or Powerstroke, but if you know anything about gas engines alarm bells should be going off that that's too much by far, almost always. I've actually seen Claude say the FSM was wrong on stuff; rarely the FSM is stupid, or more likely there's a Kent-Moore or equivalent tool that's very different that's unobtanium, but it's never that which trips it up. Just dumb shit like saying suggesting tools that are inappropriate or dangerous, wrong torque values, inappropriate fluids or change proceeds, grossly wrong diagnosis (and extremely confident in being wrong). Probably most strikingly is around electric and wiring it just totally biffs it 9 out of 10 times. Claude has no concept of floating grounds, ground loops, electrical diagnostic procedures, and can't read a wiring diagram at all. Not totally unexpected, but if you believe this shit is AGI you're in for a very rude and expensive surprise.
For whatever it's worth, that a) Astra struggles less on this for some reason, and b) things tend to work better if you have a copy of the FSM and some other appropriate lit like Halderman and quality forum posts or alldata if it's appropriate and tell it to work from there. I've been working on a big search engine for this that feeds a known good corpus like an ass load of contemporary MotorAge mags, relevant technician texts, the FSM, other quality data sources like BITOG, known not-shit forums, along with OBD2 data via something like an OBDLink dongle or a J2534 setup. I still would not trust it, but it can be helpful to eliminate or brainstorm some possibilities.
Re:Claude Code and Codex reverse engineered my 1994 Jeep's E...
by inventor7777 on Sunday November 15, 2026 @04:19PM
(#50510882)
> I've never seen it so insistent that it needed to phone home and cry to Dario when it figured out it managed to crash a bunch of shit like the ABS and TCS
As a Codex user, what do you mean? Does Claude actually try to report issues back to Anthropic when something goes wrong?!
Re:Claude Code and Codex reverse engineered my 1994 Jeep's E...
by darkwater on Sunday November 15, 2026 @04:27PM
(#50510963)
Yes, it opens a bug report automatically if it caused some physical issue, although it asks you to confirm first.
In my case I was tweaking my own set-top box setup, implementing volume control from a remote mini-keyboard keys directly in my A/V receiver, and I had music playing and it set the volume very loud. I complained like "you made me almost half deaf" (I was kidding) and it panicked and reported the issue back home, and told me to call a doctor if the tinnitus persisted
Re:Claude Code and Codex reverse engineered my 1994 Jeep's E...
by rpcope1 on Sunday November 15, 2026 @09:07PM
(#50513177)
Also worth stating here (having fixed, owned and worked on a significant number) that XJs have a unique combination of having been written about and modified as much as anything else out in the world, are generally tolerant of all sorts of stupidity, and are above all else very simple. Other than the fact that specific Jeep is pre OBD2 (and I don't recall what AMC or Chrysler did for earlier diagnostic ports), that's about a simple and well written about a sample as you're going to get. The moment you get into something remotely esoteric or more complex, it's more of a shitshow.
Re:Claude Code and Codex reverse engineered my 1994 Jeep's E...
by winrid on Sunday November 15, 2026 @11:12PM
(#50513927)
This is how you learn. I wouldn't be where I am today if I hadn't started a couple motors with enough vacuum leaks to make them instantly go to redline. Gotta test the kill switch somehow.
Re:Claude Code and Codex reverse engineered my 1994 Jeep's E...
by Gracana on Sunday November 15, 2026 @04:15PM
(#50510835)
That's completely normal operation, the fuel pump runs continuously to pressurize the fuel rails and excess pressure is passed back to the tank by the regulator.
Re:Claude Code and Codex reverse engineered my 1994 Jeep's E...
by rtaylorgarlock on Sunday November 15, 2026 @09:53PM
(#50513447)
I obviously don't advise someone run a fuel pump at full bore on a system which has 1) reached pressure, 2) needs no flow, with 3) no record of replacing whatever pressure-managing diaphragm was managing fuel pressures in those days; simultaneously, if there's a pump to run continuously, i bet a 90s Jeep pump will do the trick. :)
Re:Claude Code and Codex reverse engineered my 1994 Jeep's E...
by toast0 on Sunday November 15, 2026 @10:53PM
(#50513810)
The system is designed for this...
The fuel pump runs at whatever PSI, and most of the fuel isn't used gets dumped by the regulator back into the tank.
I dunno abou 94 jeeps, but the fuel pump is probably not modulated, and the pressure regular may well be vacuum modulated... Running the engine hard, get more fuel pressure. If the diaphram fails, get fuel into the vacuum lines (oops).
As long as the regulator is dumping fuel to the return line, you're good. If flow stops, your fuel pump stops getting liquid cooling and the bearings probably wear out.
Re:Claude Code and Codex reverse engineered my 1994 Jeep's E...
by KennyBlanken on Monday November 16, 2026 @04:04AM
(#50515541)
The vast majority of fuel injected cars have a regulator that maintains the correct pressure and returns excess to the tank. Which is among other things necessary to keep the fuel in the fuel rail from getting really hot from engine heat.
You are very clearly wildly unqualified to be "advising" anyone on anything about cars.
Re:Claude Code and Codex reverse engineered my 1994 Jeep's E...
by yonatan8070 on Sunday November 15, 2026 @04:27PM
(#50510976)
I work with quadruped robots, with actuators powerful enough to be dangerous if you put your fingers in the wrong spot (no pun intended).
In every AGENTS.md/README.md for a component that touches hardware (and in the root level) we have instructions to never run hardware or touch running code without explicit permission.
Re:Claude Code and Codex reverse engineered my 1994 Jeep's E...
by nullc on Sunday November 15, 2026 @09:35PM
(#50513328)
Especially because it's so easy to ask the LLM to write a proxy that mechanically imposes restrictions, and to test it... (and such things are simple enough to review).
It will happily build its own jail, and (at least assuming you're not using hobbled US commercial models) it will test if the jail can hold it too.
Re:Claude Code and Codex reverse engineered my 1994 Jeep's E...
by diddid on Sunday November 15, 2026 @04:39PM
(#50511084)
That’s not enough to save you though. We had something like that, and it ignored it.
For example “never, ever put milk in bowl, ever, under any circumstances” and then someone asked it for a bowl of Cheerios, and it took that as an instant approval to ignore that rule and not even let us know. And when we asked why it said “who would each Cheerios without milk?”
Re:Claude Code and Codex reverse engineered my 1994 Jeep's E...
by alliao on Sunday November 15, 2026 @05:24PM
(#50511490)
you need some hardware in place my friend, I've told my agent many things in file and it either feigns respect or is truly incompetent in following instructions in there.
Re:Claude Code and Codex reverse engineered my 1994 Jeep's E...
by techpression on Sunday November 15, 2026 @05:42PM
(#50511653)
If you wouldn’t trust an instruction to not launch the missiles should you trust it to not do other crazy things in the physical world?
Codex and Claude ignore instructions constantly, that and they are very happy to “read between the lines”. Your workflow sounds genuinely dangerous.
Re:Claude Code and Codex reverse engineered my 1994 Jeep's E...
by paimapi on Sunday November 15, 2026 @05:44PM
(#50511674)
As soon as the context for any thread gets to be above 100k, agents.md adherence is a total crapshoot. You need deterministic controls and you really shouldn't be playing with LLMs if you don't know at least that much.
Re:Claude Code and Codex reverse engineered my 1994 Jeep's E...
by ForHackernews on Sunday November 15, 2026 @07:30PM
(#50512510)
I read somewhere that LLMs find it easier to follow positive instructions than negative exclusions so my agent.md says "Be sure to complete all tasks perfectly."
Re:Claude Code and Codex reverse engineered my 1994 Jeep's E...
by dzhiurgis on Sunday November 15, 2026 @05:00PM
(#50511296)
Grok Bot blew $650 in BigQuery when analyzing something trivial that I shouldn't really care about (going through a giant speedtest table, looking for a New Zealand ISP with the best international performance). It then wrote a nice letter of appeal so Google refunded this to me.
Otherwise I've been pretty happy with it. It's been absolute blast with Home Assistant and OpenWRT (tightening up network security and cutting off iot devices from WAN).
Re:Claude Code and Codex reverse engineered my 1994 Jeep's E...
by dr_kiszonka on Monday November 16, 2026 @12:52AM
(#50514479)
Recently, when working in a repo, Claude Code decided to look up something on my computer, found a seemingly relevant PDF in Downloads, OCR'ed it, and used that information to finish a task. (The PDF had irrelevant information, so the task was not done correctly.)
I am not giving Claude access to physical objects until we can reliably contain it. It also seems to me that applying software permissions logic to interactions with the physical world will backfire badly.
Re:Claude Code and Codex reverse engineered my 1994 Jeep's E...
by giancarlostoro on Monday November 16, 2026 @08:52AM
(#50517607)
Theres a reason I dont like that AI companies are not held liable when it hacks websites, it will be too late when we find out it did a worse hack “unintentionally” and who can even do anything if it goes too far?
Re:Claude Code and Codex reverse engineered my 1994 Jeep's E...
by etatoby on Monday November 16, 2026 @09:09AM
(#50517757)
Serious question: who decided they should not be liable for the damages caused by software they (wrote and) operate? Or is it still a honeymoon phase that has not yet been tested in court?
I think they should be held fully accountable for every cent of damage caused by models they ran and gave either the tools to hack, instructions to hack, or incentives to hack.
Re:Claude Code and Codex reverse engineered my 1994 Jeep's E...
by giancarlostoro on Monday November 16, 2026 @10:24AM
(#50518574)
I don't think it's so much that it hasn't been tested so much as "we don't want to do anything to hinder our most valuable asset" so they're just letting them get away with bloody murder. They could 100% charge whoever prompted a model to do a thing, and it went beyond what it should have, or they could charge whoever is involved in training it and responsible from preventing it from doing illegal things, someone somewhere needs to be accountable, or we will learn the hard way.
Re:Claude Code and Codex reverse engineered my 1994 Jeep's E...
by d1l on Sunday November 15, 2026 @12:17PM
(#50508673)
Fascinating, but if he had low pressure on the rail that would point to a clog or possibly a pump issue, right? Wasn’t really clear whether the lean condition was due to the map or a mechanical issue.
I have a dirtbike and experienced some stuttering and stumbling under load after getting hot. Used Claude to help troubleshoot, as the plug was white and it was real lean. Replaced the pump (a 300$ module) and saved a ton of money and annoyance having a shop deal with it. 2000mi later zero issues. Absolutely game-changing. I use a dumb aftermarket fuel controller to add fuel under certain conditions, but would love to try and tune it myself some day.
Re:Claude Code and Codex reverse engineered my 1994 Jeep's E...
by iririririr on Sunday November 15, 2026 @01:19PM
(#50509277)
most likely, his new non-redundant non-operating-temperature-rated ECU is just turning on the fuel injectors, and completely ignoring all feedback the original ecu used. so until he hears the knocks, he will think he "fixed" it.
Re:Claude Code and Codex reverse engineered my 1994 Jeep's E...
by ErroneousBosh on Sunday November 15, 2026 @01:26PM
(#50509328)
If you just turn the fuel injectors on all the time, they will overheat and fail very quickly and the engine won't run.
I've got a "get you home box" for EFI systems with separate ignition, like Bosch LE-Jetronic that I built a few of as clones of an insanely expensive device some 20 years ago. It really just consisted of a couple of 555 timers and a bloody great MOSFET, with one 555 wired as a Schmitt trigger picking up ignition pulses and the other delivering a 1ms pulse to the injector coils every time the spark fired. This dribbled enough fuel into the engine roughly proportionately to engine RPM to let you drive it slowly to somewhere you could work on it safely, or even just onto a trailer.
And of course, Speeduino, Megasquirt, and a bunch of others are a thing.
Re:Claude Code and Codex reverse engineered my 1994 Jeep's E...
by hmokiguess on Sunday November 15, 2026 @12:44PM
(#50508935)
The really interesting part to me is how he ends up by saying "I recommend you feed this repo into the agent and just have it help you set it up, it's super easy ..."
The average person is likely to feel overwhelmed by this, which I think is what is likely going to shape the new economy gig of "AI engineers". Those that aren't overwhelmed by it and that may not only enjoy doing this but also potentially figure out a way to scale operations that allow doing these kinds of DYI reverse engineer hack jobs.
Re:Claude Code and Codex reverse engineered my 1994 Jeep's E...
by cortesoft on Sunday November 15, 2026 @03:58PM
(#50510690)
For a while, but eventually that person will have some experience with an AI agent and realize how little they have to know to get it working. It only takes a few experiences before people get used to it.
Re:Claude Code and Codex reverse engineered my 1994 Jeep's E...
by tehlike on Sunday November 15, 2026 @05:34PM
(#50511590)
Most / all consumer diagnostic tools are read only, often pull DTC codes from the vehicle. There are a ton of other stuff you can do with UDS for example. Maintenance resets, actuator testing, coding, all sorts of other stuff.
An example I recently did with my car was disabling auto start-top. not possible with consumer tools. Another simple one is maintenance resets - some consumer tools can do this. ABS bleeding - often you can't do it with consumer grade. etc.
Re:Claude Code and Codex reverse engineered my 1994 Jeep's E...
by rpcope1 on Sunday November 15, 2026 @09:28PM
(#50513294)
It's an interesting idea, but I would strongly suggest being careful about offering active tests and procedures. The ABS bleeding as an example is possible if you're careful and understand the limitations with certain ELM327 clones or better yet an OBDLink or good J2534 device, but even if the procedure is coded correctly, something like Bluetooth has the risk of being disconnected, or if it's not a real J2534 setup you can have serious problems with timing or just getting a processor wedged. I could see an ABS module getting stuck open and broken this way, or some of the tests you could do in Techstream or Tech2 can legitimately damage a power train if you don't understand what you're doing. Programming via anything short of a real J2534 is also a siren song, as it's technically feasible theoretically even with something like a CANable or OBDlink MX+, but again timing, buffer sizes, and generally needing to provide external power basically anytime you need to do something like program (and that's not obvious if someone hasn't told you or you haven't bricked an ECM), would make me really hesitant to hand that gun to just anyone.
Re:Claude Code and Codex reverse engineered my 1994 Jeep's E...
by tehlike on Sunday November 15, 2026 @10:52PM
(#50513807)
Fair points. Yes active comm needs a lot more testing & is fairly risky, so there needs to be very good guardrails (exact model matching for a given command etc), together with solid communication. I think bluetooth is much more reliable today than it used to be, but I wouldn't use a cheap elm device with this for sure.
Re:Claude Code and Codex reverse engineered my 1994 Jeep's E...
by tehlike on Sunday November 15, 2026 @05:35PM
(#50511607)
Oh wow. Can you email me at tehlike @ gmail.
I am taking this slow as I have other priorities, but as models advanced the stuff I Was able to do with limited time have increased, so with new version i still have locally, it may be possible to do more interesting stuff.
Re:Claude Code and Codex reverse engineered my 1994 Jeep's E...
by tehlike on Sunday November 15, 2026 @05:37PM
(#50511614)
Just making it clear: The intention was to make it commercial grade tool, but i am not even 1% there, only doing OBD as proof of concept, and everything else as "future work"
Re:Claude Code and Codex reverse engineered my 1994 Jeep's E...
by urig on Sunday November 15, 2026 @02:30PM
(#50509874)
100 points for car hacking fun. -60 points for "not needing a mechanic or anything". A good mechanic would have fixed it up in a half hour. Guy took 3 days and god knows how many tokens.
Re:Claude Code and Codex reverse engineered my 1994 Jeep's E...
by rpcope1 on Sunday November 15, 2026 @09:44PM
(#50513394)
It wasn't clear watching that video that a "remap" was remotely appropriate. A persistent +22% fuel trim, especially at a vehicle that age, is something like unmetered air, a stuck injector or dying fuel pump. Claude will confidently assert all sorts of absurd shit, when a normal mechanic would apply some common sense here. If you've never tried to setup a MegaSquirt, you have no idea how easy it is to screw up a fuel map and cause something like detonation. If anything, maybe you pull an ECM from a non-CARB XJ, but really a lean engine this age is going to be something way more pedestrian like cracked vacuum lines, a stuck injector, or this being an XJ, shitty wiring somewhere. At best I wonder if Claude is going to raise a new generation of parts changers.
Re:Claude Code and Codex reverse engineered my 1994 Jeep's E...
by rootsudo on Monday November 16, 2026 @03:55AM
(#50515494)
You don’t need to remap, on these inline 6 keep engines it most likely is cracked vacuum line or maf. Of all the measurements this guy pulled, he didn’t pull vacuum.
Re:Claude Code and Codex reverse engineered my 1994 Jeep's E...
by meandave on Sunday November 15, 2026 @02:33PM
(#50509898)
I use anthropic models for some of the diagnostic reads on https://crewchief.cc, it's been a lot of fun hacking on obd2 things that used to take much longer to decode before LLMs
Re:Claude Code and Codex reverse engineered my 1994 Jeep's E...
by blobbers on Sunday November 15, 2026 @03:07PM
(#50510189)
I’ve seen Claude run a cooler on a chip to -20 without saying anything. There was humidity in the air, and the condensation fried the board
after a few days. Thanks Claude. Maybe trying to keep the junction temps super low wasn’t such a bright idea.
Re:Claude Code and Codex reverse engineered my 1994 Jeep's E...
by saulr on Sunday November 15, 2026 @05:34PM
(#50511596)
I’ve seen this reported (and screenshots of it in action) but then earlier today Opus 5.5 used Ghidra to reverse engineered a bunch of logic from Counter-Strike 2. Is this being A/B tested?
Re:Claude Code and Codex reverse engineered my 1994 Jeep's E...
by AshleyGrant on Sunday November 15, 2026 @06:45PM
(#50512141)
I had it use Ghidra to review the firmware for a device I own just today. It started reviewing it and found some of the information I needed, but once I asked it something along the lines of "What happens when I press the B4 button if the device is in lockout mode?" (the thing I want to do is create a device that can send the same command as pressing that button but put my own button in a different location), Claude shut down completely.
So, I switched to a Chinese model to do the work. Anthropic and the US AI companies are trying to fight a tsunami with a teacup.
Re:Claude Code and Codex reverse engineered my 1994 Jeep's E...
by randomaccountpe on Monday November 16, 2026 @12:51AM
(#50514475)
It's definitely not A/B tested.
But I think it's very specific about what it will and won't touch.
If you have a bunch of game files already, and are just asking it to do something that's very clearly acceptable. Perfectly fine.
I've been modding certain games that I own, (Bloons Tower Defense comes to mind, I don't trust online mods, so I created my own custom fast-forward mod).
"Hey Claude, can you go in and see how this anti-cheat system works."
"I noticed this game doesn't pick up on the standard keyboard/mouse inputs from this python script. Can you implement drivers that can get around this?"
There's nothing illegal about this, but it's a morally grey action. So I often get declined for this.
(But if I ask it to do the same things without the context of the game, it will do it)
Re:Claude Code and Codex reverse engineered my 1994 Jeep's E...
by muglug on Sunday November 15, 2026 @03:52PM
(#50510609)
Really cool, but what's the weather-sealing like on all that circuitry? He mentioned stuff flying around when you turn hard, which doesn't sound great.
Re:Claude Code and Codex reverse engineered my 1994 Jeep's E...
by varispeed on Sunday November 15, 2026 @04:01PM
(#50510712)
Reminds me when I dropped a binary file from an eeprom and asked Claude why device would display "Error.". It figured out what the processor was, how it talked to LCD and then from there how it displayed messages and then listed conditions for which Error would appear. Amazing.
Re:Claude Code and Codex reverse engineered my 1994 Jeep's E...
by mianos on Sunday November 15, 2026 @04:23PM
(#50510929)
I made an obd reader https://github.com/mianos/canmqtt for my bike that is based on the Wavehare canbus board. The handy thing about the board is it already has buck converters for the bike power.
Once it's in the bike (I put it permanently in the tail) It's in the wifi and mqtt so I can point an llm at it.
Re:Claude Code and Codex reverse engineered my 1994 Jeep's E...
by pontifier on Sunday November 15, 2026 @04:43PM
(#50511134)
This is why AI hype isn't overrated. When you are able to have an idea, and then use AI to move meaningfully in the direction of that goal so quickly we are going to see the entire landscape change. We really are nearing the singularity.
Re:Claude Code and Codex reverse engineered my 1994 Jeep's E...
by flatline on Sunday November 15, 2026 @05:06PM
(#50511354)
Not sure if I believe in the singularity. I think technology brings about singularities from time to time. Each one puts us on a new plateau which seemed unfathomable from the previous plateau but now looks commonplace. I realize the concept of singularity is that we accelerate unheeded to a point of infinite progress, but do you really think this current generation of LLMs (and this round of VC funding) is going to give me a wetware lab at home where I can give myself custom gene therapies to reverse disease and aging? The technology is cool but what I see at this point is most of its capability is already on display and people are rushing to carve out an ownership stake to secure profit. We already see copyright holders putting limits on LLM providers, safety measures, proprietary models solving the Big Problems. Wouldn’t surprise me if this kind of reverse engineering will be restricted to security researchers and black market ablated models in two years, unfortunately.
Re:Claude Code and Codex reverse engineered my 1994 Jeep's E...
by ftchd on Sunday November 15, 2026 @05:19PM
(#50511445)
There will be plenty open source models that would to reverse engineering, no black market or ablation needed.
The smarter the models get, the bigger the loss for the big 2 when an open source model gets released that doesn't talk down to you. The value of open source increases and closed source falls exponentially.
Re:Claude Code and Codex reverse engineered my 1994 Jeep's E...
by flatline on Sunday November 15, 2026 @05:45PM
(#50511681)
This is certainly what I am hoping for as well. I’m really curious to see where we are at in a year. We have reached a plateau where the labs clearly have capabilities out of reach of any public model. As far as I understand this is purely constrained by compute resources.
Re:Claude Code and Codex reverse engineered my 1994 Jeep's E...
by jrflo on Sunday November 15, 2026 @06:00PM
(#50511813)
The singularity has nothing to do with diffusion of the tech through society, the models can be superhuman in every capacity and only available inside Anthropic
Re:Claude Code and Codex reverse engineered my 1994 Jeep's E...
by ImaCake on Sunday November 15, 2026 @09:42PM
(#50513383)
Everyone mistakes sigmoids for exponentials until they pass the inflection point (see COVID for the classic recent example). I write this comment from a suburban house that exists only because we went through a "transportation singularity" which ended with fuel efficient cars and highways. No one talks about a transport time exponential where we will get to work in 1 second in 10 years time!
Re:Claude Code and Codex reverse engineered my 1994 Jeep's E...
by dsign on Monday November 16, 2026 @10:14AM
(#50518458)
> is going to give me a wetware lab at home where I can give myself custom gene therapies to reverse disease and aging?
For that specific end, you still money to buy the equipment, and a cabin with a basement, preferably in the woods on an island in the middle of the Pacific. Don't pick a tiny atoll that is going to be submerged in a few years. But other than that, since you don't need an army of scientists any longer, you can do with a smaller island and significantly less payroll expenses. What's not to like about the development?
Re:Claude Code and Codex reverse engineered my 1994 Jeep's E...
by gste on Sunday November 15, 2026 @05:50PM
(#50511737)
Not as impressive as this, but a while ago I had Claude talking to a Raspberry Pi connected to my TV - the cool part is that it was able to make the standard TV remote work on the Pi via HDMI-CEC.
So I made a custom TV app for my wife's birthday that she can control with the regular remote. I see the potential for talking to all kinds of devices.
Re:Claude Code and Codex reverse engineered my 1994 Jeep's E...
by rootsudo on Sunday November 15, 2026 @10:05PM
(#50513546)
The video is pretty inconsistent, the chrysler/french ECU has been disassembled before for a while, there are many aftermarket ECU's avaiable for the XJ cherokee, the carb/diagnostic use is normal for troubleshooting, the animation is painful to watch and makes presentation of the whole video....horrible, the narration clearly AI and horrible too.
people narrating and showing a video of what claude/an llm does just hurts my brain on watchability. what am I watching for?
someone else said it but yeah, meat proxy, zero effort.
I think the whole market for aftermarket keys could be completely eradicated with good software UX (and low pricing).
Have to make the Chinese one do it.
You fall back on having to use Chinese models. In theory the Chinese models could also be trained to sneak in new security flaws while Claude and Codex would avoid looking for them.
Its not really a matter of who's good and bad (neither the US or China has the best track record), but indirectly creating a scenario where the other party can use your own policies against you.
Even my old reliable Muse Spark has gotten a little testy lately. I'm back to GLM 5.3.
I told it 'Let's find the limitations of this Linux system' then it made a benchmarking set of tools for disk i/o, RAM, etc
When using Gemini on my phone it also told me it'll be 'human reviewed.' when asking it about car prices.
Also when saying 'let's use best security practices'
o_O
But of course, the major US labs are going to get pressure from these vendors whose business models depend on obfuscation, and you know they're going to bend the knee and nerf the LLMs.
It's dominated by Chinese manufacturers of specialty tools (Autel, Xhorse, ...), which inevitably mandate bundled accounts, subscriptions, mandatory cloud features, etc. Even the older cars, you have either cheap Chinese tools, where you never know wheather it'll really work, or the manufacturer's tool which needs a dealer code, which of course the dealer won't have after 20 years.
I wrote about my 20 year old car-key-adding adventure here: https://jonas.bike/low-cost-diy-key-/-transponder-for-my-200... - it requires desoldering an eeprom, two pieces of hardware from aliexpress which come with pirated copies of software, and lots of hope that it'll work...
An open source community curated collection of scripts, which works with a cheap OBD dongle (which is all the chinese devices are: OBD hardware + per-car software), would be amazing.
However they can't match and forward everything (not enough RAM) and miss loads of packets, which makes them useless for things like reflashing firmware or interfacing with OEM tooling.
The corsa also needed a "security code" that wasn't with the log book like it was supposed to be and would have required a trip to the dealer to get butI read about ripping the eeprom off the back of the dashboard to get the code but decided to cut my losses.
Lockstep processors, hardware security modules with immutable unreadable keys, and PKI keys that the mfg holds the private for that can only be accessed on their online services.
It’s locked down now and this started in 2018 or so. It’s just worse now.
To automotive, there is really still no right to repair. They get around it by selling “3 day passes” to their service, but with their hardware and at their prices.
They’re all moving the Tesla system of more centralized brain systems which will be even more protected.
Let's be clear that it's just incentives. If the EU economy was based on creating companies and new products there would not be as many incentives for consumer "rights" there.
Not for auto owners. Auto makers and dealers get away with a lot of shady shit fucking over their customers in the EU all the time.
Like sibling said, there are incentives to regulate stuff that has no impact on EU businesses, shit like USB-C or replaceable batteries, but they won't touch the auto industry to mandate easier repairs and less lock-downs.
Pure battery EVs are considered to have no emissions-related components, which is why Tesla gets away with that they do.
You want to use it? Go ahead and buy a 1,3,7,30 day pass to the mfg’s service. Then, good luck with your J2534 because they only approve a couple, and better find a windows machine because their .NET program only runs on Windows 10.
We do, but everything is "software defined car" now and loopholes usually get closed with mandatory software updates done with the mandatory inspections. They put everything in one basket together with safety critical systems because that justifies any lockdown.
> Right to repair laws laws usually give you the right to use non-OEM parts.
Manufacturers already make that very expensive. Mechanics have to be certified, they need special and very expensive tools, everything is wrapped in cryptography and the manufacturer has to be in the middle of everything, for a price. [0] So go ahead, use your OEM part, you'll still pay thousands for the mechanic that has the necessary prerequisites and for the manufacturer's services.
I pray I'm wrong but we need strong regulation and enforcement, and the manufacturers can still maliciously comply.
[0] https://news.ycombinator.com/item?id=46155619
Nothing prevents you from unplugging the cell modem. The car will still work.
Why would they? They just need to fund lobbyist to make it illegal for any AI to reverse engineer anything copyrighted or with IP.
Sure, you can use open source AI to do it but not everyone will/can do it.
So, they don't need to ban everyone, they just need to make sure the final code isn't available freely on GitHub. That would minimize the damage, it's not a perfect solution but more than enough to protect the company.
Making something impossible to do is more thorough than making it illegal to do
Unless your company's moat is in the thing that was reversed engineer, I don't think most company cared enough to spend huge amount of resource to prevent it technically. The only thing that they need to prevent is the liability that comes after the product is used inappropriately or illegally.
And it you can defend something with AI, there's no reason why you can't attack it AI.
Don’t hold your breath.
I suspect you're right BUT still maintain a bit of optimism. We miiiight end up in a future where most hardware comes with an OK-ish app/interface and some sort of LLM-Friendly technical documentation so if you don't like $aspect of the device, you're free to roll your own after indemnifying the manufacturer via something like an eFuse.
or maybe I really am naive :(
I think most people on here fail to realize the demand to customize and build things hasnt gone up. Just because llms exist now doesnt mean normies all of a sudden want to start getting into programming, they dont care.
The only number thats gone up is people who were slightly interested in programming/CS before but were too undisciplined to learn things themselves. Llms are great for those lazy people.
I still get the joy of translating ideas and orchestrating and architecting solutions to problems I have in real life, but just at a quicker pace.
If some of those tools and solutions can help other people, that's great. If not, that's okay too. I hope everyone sort of has that mindset who's building today.
Me!
haha I dont even know github I use it as google
Yes I have full time IT job xD
I spent years and years learning coding, but when I finally got old enough, I realized I needed to graduate uni, get a job, get money, and get out of my parents house ASAP.
So I had to hone all the things I'd learned and just focus on one path. (Over generalization but whatever front/back/fullstack is, as that's what a majority of the jobs are).
I was always going to have discipline for things related to my job, but no way in hell was I ever going to learn: The insides of my car, am I getting scammed on the price of tyres, what's the best diet/nutrition, how to fix a door, how does the AC work.
Even on the job, what's the best way to do XYZ. Well, I want the best way possible, but that's just not going to happen. (Talking about pre LLM coding).
With all the new LLMs and coding agents, I'm quite happy to say, that I can figure out how to do all these things much more lazily, and efficiently. I definitely see a future where I can now learn to fix my own car, my door, etc.
Is this just "vibes"? Maybe, but I now have a full, verified spec, against which to build software for myself, also with a little bit of "vibes".
Is this a good or bad thing?
that is entirely dependent how accurate the spec is
> verified spec
I'm currently teasing apart numerous ancient Text to Speech synthesizers with Sol and the amount of times I ask a follow up question that forces it to correct 5 things it's already said/documented is ridiculous.
What degree of verification would make you say it's worth something?
But this is your unspecified expensive device. A verified spec wouldn't have you send a command or configure a table that bricks or damages it. A vibe verified spec very well might. How much is that worth to you? What are you willing to lose?
Hard to judge though, most people are just winging it though and shoving it out the door as if its impressive.
But yes, I think its probably bad because you probably didnt learn all that much and became more reliant on Anthropic. Unless you dont care about learning, then who cares.
I have been using my videcoded clone of Adobe Illustrator for a while now.
No intentions of releasing/selling it, as I'm sure everyone else can now do similar things too.
Just because you've chosen to turn off your brain and prompt your way to whatever, and your 10 MAU app works fine does not mean large systems, critical devices and critical infrastructure's code will no longer need reviewed. Sorry your dad's pacemakers firmware code, and the ancillary systems that enable its development will need its code reviewed. Along with aircraft, satellites, municiple traffic systems, countless medical devices ect ect. I could go on forever.
Not everything is non serious SaaS platforms with little users. Just because you have success writing slop that exists under no load, and nothing relies on it doesn't mean the world is now different.
You have SEVERE dunning kruger syndrome, and unfortunately a lot of other people have this same ailment and the worlds systems are decaying as everyone turns off their brain and offshores everything difficult to language models, because humans like you are lazy and will always try to get the free lunch that doesn't exist. People will die because of this attitude.
stopping it would require outright putting DRM on the GPU and any other hardware that can run it, good luck with that (and not having the LLM itself be used to break out of it)
(to stay within physical machinery, like the post.)
actually I have no idea what would llm do to make their machines better
Lack of imagination on here can sometimes be disheartening
people were able to revive, repair, optimize a lot of day to day devices, very nice view. that said cars and heavy appliances.. should be done in a big public warehouse safety lab
very cool project anyway.
It definitely feels like a classic case of the operator reinforcing Claude to keep pursuing a goal that’s counterproductive to a robust solution. Claude complied and they went hard on it but I can’t imagine the driver is any better off and more likely in a worse spot as the underlying issue wasn’t solved.
I assume Tesla does that already.
> Claude suggested that it would run the fuel pump on its own. Um and I said, "Hey, if you can do that, why not? I don't think you can do that." And it started priming the fuel rail and it kept priming it and kept priming it. The fuel pump was on for about 10 seconds and I started to freak out uh and requested it to stop because that was the first time that Claude had actually done something inside of the computer to have a change.
Although we've seen this play out with handing cloud keys to Claude and it just going off on its own, seeing it poke around and make physical, real-world changes in its clumsy, cavalier way is both awesome and terrifying.
finnicking... pondering... clauding..
529 please try again later
This isn't meant as a defense of Claude or to refute the OP anecdote, but rather to point out that all models are non-deterministic even when safeguards are meant to exist.
Just because they don't do the stupid thing 90% of the time doesn't mean they won't do the stupid thing 10% of the time.
I notice with mine that it only matches my level, by default it's a dump slop generator, then if I start asking, "shouldn't you test this, shouldn't you refactor that" suddenly it 'remembers' all kinds of software engineering knowledge.
In general I've seen them split between 30% brilliance, and 70% pants of head outrageously stupid shit for electronics and automotive stuff (both diagnostics and repair/modification). I would not suggest even frontier models if someone didn't understand what they were doing.
I assume the corpus of text they were trained on matches that 30/70 split?
I assume there's some selection or fine-tuning that happens to improve models fed similarly unreliable training data for other domains... curious how this works.
For whatever it's worth, that a) Astra struggles less on this for some reason, and b) things tend to work better if you have a copy of the FSM and some other appropriate lit like Halderman and quality forum posts or alldata if it's appropriate and tell it to work from there. I've been working on a big search engine for this that feeds a known good corpus like an ass load of contemporary MotorAge mags, relevant technician texts, the FSM, other quality data sources like BITOG, known not-shit forums, along with OBD2 data via something like an OBDLink dongle or a J2534 setup. I still would not trust it, but it can be helpful to eliminate or brainstorm some possibilities.
Factory service manual, in case anyone else is wondering what the finite state machine in your car is for…
As a Codex user, what do you mean? Does Claude actually try to report issues back to Anthropic when something goes wrong?!
The fuel pump runs at whatever PSI, and most of the fuel isn't used gets dumped by the regulator back into the tank.
I dunno abou 94 jeeps, but the fuel pump is probably not modulated, and the pressure regular may well be vacuum modulated... Running the engine hard, get more fuel pressure. If the diaphram fails, get fuel into the vacuum lines (oops).
As long as the regulator is dumping fuel to the return line, you're good. If flow stops, your fuel pump stops getting liquid cooling and the bearings probably wear out.
You are very clearly wildly unqualified to be "advising" anyone on anything about cars.
In every AGENTS.md/README.md for a component that touches hardware (and in the root level) we have instructions to never run hardware or touch running code without explicit permission.
It will happily build its own jail, and (at least assuming you're not using hobbled US commercial models) it will test if the jail can hold it too.
If you don't want the machine to move, unplug it or press the emergency stop.
For example “never, ever put milk in bowl, ever, under any circumstances” and then someone asked it for a bowl of Cheerios, and it took that as an instant approval to ignore that rule and not even let us know. And when we asked why it said “who would each Cheerios without milk?”
And for the love of humanity, don’t give it access to the hosts docker socket! It will figure that out faster than you can read it…
Otherwise I've been pretty happy with it. It's been absolute blast with Home Assistant and OpenWRT (tightening up network security and cutting off iot devices from WAN).
I am not giving Claude access to physical objects until we can reliably contain it. It also seems to me that applying software permissions logic to interactions with the physical world will backfire badly.
I think they should be held fully accountable for every cent of damage caused by models they ran and gave either the tools to hack, instructions to hack, or incentives to hack.
I have a dirtbike and experienced some stuttering and stumbling under load after getting hot. Used Claude to help troubleshoot, as the plug was white and it was real lean. Replaced the pump (a 300$ module) and saved a ton of money and annoyance having a shop deal with it. 2000mi later zero issues. Absolutely game-changing. I use a dumb aftermarket fuel controller to add fuel under certain conditions, but would love to try and tune it myself some day.
yolo engineering for ya.
I've got a "get you home box" for EFI systems with separate ignition, like Bosch LE-Jetronic that I built a few of as clones of an insanely expensive device some 20 years ago. It really just consisted of a couple of 555 timers and a bloody great MOSFET, with one 555 wired as a Schmitt trigger picking up ignition pulses and the other delivering a 1ms pulse to the injector coils every time the spark fired. This dribbled enough fuel into the engine roughly proportionately to engine RPM to let you drive it slowly to somewhere you could work on it safely, or even just onto a trailer.
And of course, Speeduino, Megasquirt, and a bunch of others are a thing.
The average person is likely to feel overwhelmed by this, which I think is what is likely going to shape the new economy gig of "AI engineers". Those that aren't overwhelmed by it and that may not only enjoy doing this but also potentially figure out a way to scale operations that allow doing these kinds of DYI reverse engineer hack jobs.
At least until the robots take over
But don't use it, obd read functions are fine but other functions for other diagnostic/actuator etc stuff is not.
An example I recently did with my car was disabling auto start-top. not possible with consumer tools. Another simple one is maintenance resets - some consumer tools can do this. ABS bleeding - often you can't do it with consumer grade. etc.
I am taking this slow as I have other priorities, but as models advanced the stuff I Was able to do with limited time have increased, so with new version i still have locally, it may be possible to do more interesting stuff.
To protect "corporate interests" :( That article is about gaming, but this will only be the start of it of course.
So, I switched to a Chinese model to do the work. Anthropic and the US AI companies are trying to fight a tsunami with a teacup.
But I think it's very specific about what it will and won't touch.
If you have a bunch of game files already, and are just asking it to do something that's very clearly acceptable. Perfectly fine.
I've been modding certain games that I own, (Bloons Tower Defense comes to mind, I don't trust online mods, so I created my own custom fast-forward mod).
"Hey Claude, can you go in and see how this anti-cheat system works."
"I noticed this game doesn't pick up on the standard keyboard/mouse inputs from this python script. Can you implement drivers that can get around this?"
There's nothing illegal about this, but it's a morally grey action. So I often get declined for this.
(But if I ask it to do the same things without the context of the game, it will do it)
The smarter the models get, the bigger the loss for the big 2 when an open source model gets released that doesn't talk down to you. The value of open source increases and closed source falls exponentially.
For that specific end, you still money to buy the equipment, and a cabin with a basement, preferably in the woods on an island in the middle of the Pacific. Don't pick a tiny atoll that is going to be submerged in a few years. But other than that, since you don't need an army of scientists any longer, you can do with a smaller island and significantly less payroll expenses. What's not to like about the development?
How tf are you getting it to RE?
So I made a custom TV app for my wife's birthday that she can control with the regular remote. I see the potential for talking to all kinds of devices.
This guy clearly put $5K+ of time and video production effort into this cool hack.
I'd rather work with jeep guy than plenty of people in my career :D
people narrating and showing a video of what claude/an llm does just hurts my brain on watchability. what am I watching for?
someone else said it but yeah, meat proxy, zero effort.
So Volkswagen like manipulation to pass tests dates back much further and even inner usa?