Hacker Newsnew | past | comments | ask | show | jobs | submit | mktemp-d's commentslogin

For typing “yes” or “y” automatically into command prompts without interacting, you could have utilized the command ‘yes’ and piped it into the process you’re running as a first attempt to solving the yes problem. https://man7.org/linux/man-pages/man1/yes.1.html


I don't think this is an actual problem and the prompt is there for a reason.

Piping 'yes' to command prompts just to auto-approve any change isn't really a good idea, especially when the code / script can be malicious.


And here I was hoping OP was being sarcastic. Yet it‘s reasonable we‘re nearing an AI-fueled Homer drinking bird scenario.

Some concepts people try out using AI (for lack of a more specific word) are interesting. They will add to our collective understanding of when these tools, paired with meaningful methods can be used to effectively achieve what seemed out of reach before.

Unfortunately it comes with many rediscovering insights I thought we already had, badly. Others use tools without giving consideration to what they were looking to accomplish, and how they would know if they did.


Isn't that the point of vibe coding? You don't even look at the code. Just trust the llm to take the wheel.

https://x.com/karpathy/status/1886192184808149383?lang=en


40K - 38K means 2K lines of actual code.

Which could mean that code was refactored and then built on top of. Or it could just mean that Claude had to correct itself multiple times over those 459 commits.

Does correcting your mistakes from yesterday’s ChatGPT binge episode count as progress…maybe?


If it doesn't revert the corrections, maybe it is progress?

I can easily imagine constant churn in the code because it switches between five different implementations when run five times, foing back to the first one on the sixth time and repeating the process.

I gotta ask, though, why exactly is that much code needed for what CC does?

It's a specialised wrapper.


How many lines of code are they allowed to use for it, and why have we put you in charge of deciding how much code they're allowed to use? There's probably a bit more to it than just:

    #!/usr/bin/env bash
    while true; do
      printf "> "
      read -r USER_INPUT || exit 0
      RESPONSE=$(curl -s https://api.openai.com/v1/chat/completions \
        -H "Authorization: Bearer $OPENAI_API_KEY" \
        -H "Content-Type: application/json" \
        -d "{
          \"model\": \"gpt-5.2\",
          \"messages\": [
            {\"role\": \"user\", \"content\": \"$USER_INPUT\"}
          ]
        }")
      echo "$RESPONSE" | jq -r '.choices[0].message.content'
    done


> How many lines of code are they allowed to use for it, and why have we put you in charge of deciding how much code they're allowed to use?

That's an awfully presumptious tone to take :-)

I'm not deciding "This is how many lines they are allowed", I'm trying to get an idea of exactly what sort of functionality that CC provides requires that sort of volume.

I mean, it's a high-level language being used, it's pulling in a lot of dependencies, etc. It literally is glue code.

Bearing in mind that it appears to be (at this point anyway) purely vibe-coded, I am wondering just how much of the code is dead weight - generated by the LLM and never removed.


Your forget to insert the part where the President asks the convicted defendant if they want to finance their pardon with Klarna or Affirm in the Presidential Library's checkout page


Financed through purchases of the President's own crypto coin


Considering how popular the show was at its peak, the fact that there is still more content to come and how IPs are being rehashed all the time, it would a non zero chance that they redo the entire thing someday.


what's funny is that, from what i recall hearing, the creative forces behind the tv show essentially got bored with the show and crammed a ton of stuff into season 8 and kinda just phoned it in. iirc hbo was willing for one or two more seasons after 8


The producers got like a 100 million deal from Disney to start making shows for them, IIRC.


that's an insane amount of money if it's accurate. wonder why they couldn't just negotiate with hbo to let them free and have someone else do their part.


Disney ruining shoes they not even own, thats a first..


Telegram users spinning up their own honeypots and blindly trusting a client/server message encryption system is never not a great idea for new grass root criminal enterprises.


I find that some folks who know just a little about security are some of the worst at it. Their ability to confidently make terrible choices and inexplicably expose themselves to more risk than some rando citizen is amazing. It's like their strong enthusiasm / personal beliefs drive them head long into inexplicable choices and now their eggs are all in one insecure basket and they put a lot of foolish things there.

In contrast a more nervous / unknowing person might think "oh man I better not talk about this anywhere, I don't know who could be listening".


It's like that classic bell curve troll meme - "oh man I better not talk about this anywhere, I don't know who could be listening" is a correct instinct, especially in a western country. Doing anything on the web, whether it be crimes and ecommerce, is absolutely not anonymous. They re-use handles or emails that have personally identifying information, they don't use clean workstations, they brag (dumbest opsec thing ever, giving away information for absolutely no reason than your big ego), they taunt law enforcement. Osama bin Laden basically vanished off the face of the planet when much of the world's most powerful intelligence and militaries were hunting him, and he wasn't hiding in a cave, but he was not connected to the internet in any way whatsoever and communicated via courier, which still got got. The only reason you are anonymous or think you are anonymous is because no one powerful or determined enough has gone looking yet. This is a fact I am convinced of, and I am much more fearful of successful obfuscation tactics and red herrings left on purpose rather than a 20 year old kid engaging in a fantasy that he's so l33t he'll never get caught.

The other thing I'd say to any aspiring criminals out there is it's usually much less stressful and still profitable to get gainful employment if you are actually a talented hacker. Most of these guys seem like script kiddies, that do not understand the ramifications of what they are doing. Some of these breaches will be felt and cleaned up for decades, all so they could get a laugh and a few shekels and their e-peens stroked by other criminals.


>especially in a western country

I'm not really convinced this is a distinction that matters.


To me it does, in terms of data privacy and isolation if your goal is to remain undetected by anyone you wish not to be - it's the wild west, even if you believe the GDPR has been effective. Even if you do believe it's just as bad in the rest of the world, we're heading into some sort of crisis when sufficiently powerful computing becomes commercially available (if it isn't already somewhere in a lab) and all the data these countries have been hoovering up and storing for who knows how long becomes decrypted, I would much rather be living in other parts of the world in terms of my privacy if/when that day comes.


Are you under the impression things are somehow better in e.g. Saudi Arabia or Russia or China? Maybe if the qualification was "developed countries", because developing ones might not have the budget, but "the west" is just wrong.


I'm not sure I understand.

I guess I was saying that I don't see "especially" the west as far as privacy goes.


By grass root you mean not state sponsored? Agreed it’s not a good idea using Telegram as a server, people forget bots have chat history you can replay too


The engineering requirements to build a branch of a company with the security needs of specialized SOC would be crazy endeavor and to the point that whatever comes of it would probably look as foreign to the main company's function than what currently exists. Having a separate non-profit agency handle these cases would not only allow for the agency's incentives to align, but also the overall security design approach to its facilities and internal networks. It's a win-win for both private and public interest imo.


With regards to US politics, there's campaign funding considerations as to how corporations assist candidates https://defendcampaigns.org/donors


You can't just lay out a supposed fact that 99% of users don't care about speed without providing some sort of citation...


I think you misread, I read it as 99% of users don't care about anonymity


A good book to read regarding this topic is "Death in Mud Lick" by Eric Eyre who was a reporter for a local newspaper in WV during the early discovery phase of the epidemic.

The book details the methods used by the drug companies to pedal these drugs into small communities, the tragic victims, and lastly the refusal of drug suppliers to admit any wrong doing. There's also a slight dose (no pun intended) of WV AG's sheer incompetence on the matter due to donations and gifts by said drug suppliers.

Super depressing book, but great read nonetheless.


Painkiller is a great little short series on them too. does a great job at portraying all of what you mention in a six hour tv format. it can get a little on the nose sometimes but the meat is 100% true


*peddle


Fun fact: early opioid deliveries were actually, in fact, almost exclusively done by bicycle!


If they used those penny farthing bicycles I’m sure they needed quite a few painkillers.


It always baffles me how long it takes for Windows to search for a file while the GNU's `find` command churns through filenames and paths like it's nothing.


Windows Everything finds any file globally in the blink of an eye, let's you sort in real time by size/date, live-edit your query, and also perform operations on the result in the list, so much better that any find


Just so there's no confusion, eviks means "everything" by voidtools, which is really amazing. The difference in search speed is like day and night.


It's highly optimized, but it still needs quite a bit of RAM when the index is large. But RAM is cheap nowadays. It's the only thing I really, really miss when I am on Linux. (there is FSearch, but it's not quite as great, yet)


Nirsoft's Search My Files gets me pretty close to Win2000 search.


Work locks me into Outlook but god I wish I could just grep my inbox


If your sysadmins are kind, Outlook can expose IMAP/SMTP.


Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: