Reports indicate that cheaters need a full Windows reinstall to recover?
> installs malware
> malware does malware stuff
surprised_pikachu.jpg
Of course it's hard to write sympathetically about cheaters, but this is exactly the sort of nonsense you'd expect from installing a hostile kernel extension.
It doesn't need a full reinstall. DRA cheat devices are not compatible with Windows when IOMMU protection is enabled. You just have to disable the protection to get the device driver to be able to read data it shouldn't be able to again.
Can't believe I had to scroll for this. Trust the OEMs to do something very close to what enthusiasts might want, then immediately torpedo it with "but we won't sell it to you."
This level of conversion isn't exactly trivial but it also isn't rocket surgery for the kind of person who pulls an engine out for rebuild on a classic car project.
>This level of conversion isn't exactly trivial but it also isn't rocket surgery for the kind of person who pulls an engine out for rebuild on a classic car project.
If you saw the "quality" of electrical work otherwise very smart car enthusiasts do you might think otherwise.
Also look at the "DRM" controls:
"
Q. Can I increase horsepower?
a. The first-ever Chevy eCrate conversion kit has a locked system that does not allow you to increase horsepower at this time.
"
... so you're buying into a locked, digital control system, akin to what John Deere puts out.
This ranks right up there with BMW wanting to charge a monthly fee for heated seats - building in physical abilities, with digital lockouts. You know, you can buy a LS engine, and do whatever horsepower changes you want to it. For those more akin to computers than cars, this is called a "LS swap" and is common with restomods.
This is disappointing to hear and tarnishes a brand like Chevy. Fortunately, we're in a free market; I'll vote with my dollars.
Modern gas crate powertrain swaps which include engine management usually have the same restrictions; GM Connect and Cruise, Hellcrate, Ford Performance, etc. What you’re describing with LS swaps is unique to kits that come with no engine management; it would be as though this kit were sold with no inverter and motor controller. Of course, reverse engineered aftermarket tuning is more readily available for gas ECUs, but that’s just a matter of market forces, not the OEMs.
It kinda does have that effect, because large companies are allergic to AGPL by policy, even if they otherwise have processes to get GPL software into use. The very companies that have the pocketbook to pay for your software, are also structurally incapable of using it as FOSS. Smaller ones that are more agile about how they incorporate it, have less willingness to pay for a different license.
Sure, but there's a case I'm particularly aware of where one of the major cloud infrastructure providers was about to host a significant AGPL-licensed project without modifications because their lawyers had reviewed it and determined it would have been OK. The particular VC-backed open-core company then got word of it and changed their license off of AGPL. IYKYK
Here's a question: I don't understand the gap between these LLM powered voice agents vs CLI coding agents, the latter of which are obviously useful and quite resourceful at getting something done when asked in plain English.
Seems like an agent given 20-30 tool calls like "read_sms" "matter_command", and "send_email" would be able to work out what to do for things like "set the house to 72° and text Laura that I did it."
> Seems like an agent given 20-30 tool calls like "read_sms" "matter_command", and "send_email" would be able to work out what to do for things like "set the house to 72° and text Laura that I did it."
Incidentally, a major headline in the news this past week was about a coding-agent that wiped its company's entire system, including backups; which the company's staffers were confident was utterly impossible (as it didn't have any access to that system), and yet somehow, it did[1] (the TL;DR is the agent randomly came across an unprotected God-tier admin API-key/token saved to a personal text-file in a filesystem it had read-access to). If an agent can do that with only read-only access to a company's routine/everyday storage area then there's no way I'm giving it the ability to deactivate my house's fire-alarms and security-cameras via Google Home/Matter/Thread/HomeKit/X10/OhFfsNotAnotherCloudBasedAutomationScheme.
If you are really worried about that, the agent already has that access since itll go find that key anyways.
the HN thread about that case was much more of a "why are you putting your prod keys in random text files" and "the sota in prompt engineering is that putting DONT FUCKING DO THE BAD THING" makes the agent more desperate to get stuff done
putting limits at the harness level would do just fine. one LLM call, one tool call per voice message.
This is a weird comment because I feel the same about getting macOS to a useable place.
I probably have 5 or 6 things installed on my Mac like Scroll Reverse and Rectangle, just trying to beat the window manager into something that resembles useable.
I noticed this almost immediately when attempting to switch to Opus 4.6. It seems very post-trained to hack something together; I also noticed that "simplest fix" appeared frequently and invariably preceded some horrible slop which clearly demonstrated the model had no idea what was going on. The link suggests this is due to lack of research.
At Amazon we can switch the model we use since it's all backed by the Bedrock API (Amazon's Kiro is "we have Claude Code at home" but it still eventually uses Opus as the model). I suppose this means the issue isn't confined to just Claude Code. I switched back to Opus 4.5 but I guess that won't be served forever.
This is a fantastic explanation. Thank you. The only part I am not following is how it is guaranteed that 1 bit is sufficient for the error value. Is this something the Lloyd-Max algorithm is responsible for ensuring? (Seems to me that if your quantization algorithm is crappy enough, you could need a large number of bits to store the error.)
I would love to lobby to change how the law works for these cases: for some definition of "firmware" (informally "software that ships with hardware and is not intended to be selected by the consumer like a computer operating system"), add a copyright exception so that modifying the firmware in situ is treated like modifying the physical hardware, because in practice they are in fact the same thing: a single component that does a single thing.
With this, the John Deere approach to gatekeeping vehicle repair would no longer be legally protected by the DMCA or by copyright law. All the other protections afforded by copyright law would still apply: you cannot rip the firmware off the hardware and distribute it, the manufacturer is under no obligation to help you modify it, etc.
However, tools which patch or circumvent antifeatures of the firmware would now be legal to use on hardware you own: it would be legal to patch out software locks, retune engine computers, etc.
I think the law should regard the general thing being copied (firmware, software, etc) as different from a single copy. For example, the law could regard modifying a single instance of firmware similarly to the way it regards modifying a book. Right now you can take a book and mark it up or tear pages out if you want without any permission, and the only reason permission would be required for firmware is because of the ability to have telemetry and attestation. So it seems like a pretty good extension of copyright law to protect any modification of a copy but not the sale of additional copies.
Ooh this gives me an interesting passive-aggressive idea to counter pointless "is this still relevant" questions. "No, I haven't hit this in the last 2 days." "No, I haven't hit this since I gave up trying to do it with your tool." And so forth.
The less passive-aggressive version is to use this obviously-unhelpful answer of the obviously-unhelpful question, to actually have a conversation to get the PM to recognize that the default state of a ticket is in fact "no change." Ultimately that may turn into a stale bot if the PM realizes the policy they actually want is some sort of timeout, but at least it's not a time consuming meeting!
(Note, a cathartic thought experiment, but not really good manners to actually do!)
> installs malware
> malware does malware stuff
surprised_pikachu.jpg
Of course it's hard to write sympathetically about cheaters, but this is exactly the sort of nonsense you'd expect from installing a hostile kernel extension.