Christ.
81 karma · joined December 13, 2025
Christ.
One such situation:
A particularly competent "scripter" writes automation programs for the most complicated content in the game. As such, they sell at a premium. It makes you and him considerable money for many months. Eventually, the community starts to develop a free version of this same automation program. You are now in a situation where you've likely built your business around the expectation of revenue from this one "scripter" and his/her products. Maybe you've even given them a more favorable revenue share model.
Another issue that can pop up from this, especially in lower pop games, is that it's very obvious from a statistics point of view. The game developer will easily take notice if the hardest content in the game is suddenly being done many times per day by players who have never showed an affinity for such content. This in turn increases ban numbers which hurts your product's reputation.
All of this is to say: Having thousands of hours in the game you are working on is a massive benefit at every turn. It is best to work on games you genuinely enjoy.
From a regular player's perspective, if they want to kill a boss for 3 hours, they must use X amount of Y resource. If the cost of Y resource is very high, or worse, the market has not enough supply, the game becomes less fun for them.
This is addressed by bots, but is better addressed by a design solution from the developer.
All of that being said, if bots exist, the reward for going and farming that boss is going to be far less than if bots didn't exist. So still, negative.
I don't have any regret for profiting from a game being bled dry by bankers. When that stopped, so did I.
I feel that's fair.
Find call to `SetWindowsHookEx` -> Find dword (`HHOOK`) it returns -> Inject module that calls `UnhookWindowsHookEx((HHOOK)(module_base + hhook_rva));` -> profit.
Jagex has done some interesting things, but without fail, there is a blunder accompanying each.
Since then I've been using images with very simply worded prompts whenever I'm informing an agent of what is happening. Sometimes no text in the prompt at all.
It has been very very effective.
That being said, this isn't really what Karpathy was talking about. But it got me thinking a bit, and that got me to a much nicer workflow.
Most people would also agree that building a better Github is not a super easy two-person task.
What? If the person writing the article is so unfamiliar with the subject they are writing about, they likely should not be writing about it.
I'm sure there's reasons as to why. I just don't know them.
In his follow-up post he talks about him open sourcing old games as a gift, and he doesn't much care how people receive that gift, just that they do.
He doesn't acknowledge that Anthropic, OpenAI, etc, are profiting while the original authors are not.
The original authors most of the time didn't write the software to profit. But that doesn't mean they don't care if other people profit from their work.
It's odd to me that he doesn't acknowledge this.
Tested on m1 pro 2021 laptop and recent higher-end (4080, 14700k, etc) desktop. Same on both.
The fuck?
Which you said aren't being made anymore. Which is factually untrue. The best bit is, they're still being made! And there's plenty of people who are still buying them!
Why? Because a $170 pair of closed-backs sounds infinitely better than the $550 Bose Quiet Comfort Ultra nonsense.
FiiO FT1 32Ω being a prime example, if you are looking for closed back suggestions :^)
Sennheiser HD 660S2
Audeze MM-100
Meze 109 Pro
Focal Hadenys
Focal Azurys
Sennheiser HD 505
HiFiMAN Arya Stealth
Audeze LCD-5
But they don't. And won't.
In my head, I miss those days. But I don't miss lacking a moral compass. I think I miss the feeling of fewer online barriers existing.
You'd connect to a live chat (or phone call) and tell the customer support rep that it's your first day on the job and the man training you has a thick accent, making it very difficult to understand him. You then ask a hypothetical, e.g "If someone phones in asking for help with regaining access to their account, or setting up security questions, what tool do I open up?". You'd then get more specific with the hypotheticals, gradually. The only thing you cared about were the name of the tool and the steps to pull up an account by IP address. At the time, almost all ISPs had their own software suite. Verizon used something related to Coffee. I think it was just called Coffee tbh. Anyway, the goal is to get them to tell you both the name of the internal tool they use + the rough steps on how to pull up an account. Most of the time, via phone and sheer confidence, you would get the information you needed within two to three attempts. You could also take it a step further if you were bored and try to get screenshots from the rep of the software.
Edit: You'd also ask the rep for their employee identification number, if applicable. You'd then use that if you need to hard-sell it in the next call. It really didn't matter if the ID was valid, so long as it was the correct length/format. Yes, they would really just tell you. I do not know why.
After you had that information, you'd phone back, making sure you got separate rep. Depending on size of ISP, you may have had to call back after a shift change.
You inform them your regional servers are down and you're unable to connect. You could flush this out more if you had additional information on the ISPs tech. You then would go on to say "I have a customer on the line who's rightfully upset after their account was apparently accessed without authorization. The customer is saying they changed the PII on the account and they're unable to recover it." This gives you a shitty, but somewhat valid-ish excuse to pull the account up by IP. You'd then use the information attained from the first step to sell that you are indeed an employee. Name of the tool, input labels (roughly), steps needed, button names, etc. If the rep is remotely technical, hang up, try again. You'd then confirm the information on the account with the rep. It helped if you had some information about the person already, e.g first name or rough location.
Comcast was the worst offender. Charter second. Verizon was a bit more tough, but not by much. People started doing this as a first-step in targeted identity fraud, which got a lot more attention on it. Along with all the typical information (street address, postal code, state) you'd also almost always be able to get the last four of the social on the account + last 4 of any card on auto-pay.
If you're worried about this sort of thing, the best advice I can give you is to check with your ISP and see if they allow a verbal password that can be tied to your account. Anyone calling in for support or connecting to live chat would need to provide it before the account's accessed.
I'm not sure how relevant swatting is nowadays, but if you're at all in a position where you have concerns over it happening, it would be wise to phone your local police department and let them know there's a possibility this might happen. From what I remember, most of the time they ask for your cell number. In the event that this does happen, they will still send the full swat team to your residence. But they will phone your cell and ask you to come out prior to kicking down the door.
Source: was bored in when I was 15/16 and doxxed pedophiles.