[-] FiniteBanjo@programming.dev 1 points 6 hours ago

For example if a device manufacturer provides drivers for linux, or a software developer has a version for Arch, but it's missing a pkg build or config file, most users simply won't be ably to figure out how to manually install it and CORE or Flathub probably don't have any official packages for it.

There are millions of such niche cases like this every day.

[-] FiniteBanjo@programming.dev 2 points 9 hours ago

Our apologies. Let us issue a correction:

"Agents should never be allowed to program"

There.

[-] FiniteBanjo@programming.dev 11 points 9 hours ago

Let it be known that the first person to call it was actually Sam Altman when OpenAI's paper on AI Scaling Laws in 2020 subtly showed that the diminishing returns will stop showing improvement with infinite power, compute time, and data before 94% accuracy is reached.

[-] FiniteBanjo@programming.dev 5 points 9 hours ago

It will eventually be detected, but it passes tests before hitting production, that is the problem.

[-] FiniteBanjo@programming.dev 4 points 9 hours ago* (last edited 7 hours ago)

Humans generally don't hallucinate libraries or documentation. If there is a bug or error on a human maintaine repo the human in charge will generally know what went wrong and how to fix it, the AI will just gaslight your ass because the AI has no idea.

[-] FiniteBanjo@programming.dev 8 points 9 hours ago

No, humans make less mistakes. Less. That's the key here, statistical models are trained on human data so by pure logic can never, ever, under any circuimstance, reach 100% accuracy. With current understanding of LLMs with a focus on AI Scaling Laws, and more importantly of natural human language adaptation, they will never reach 94% accuracy with infinite power and infinite training. That's what the curve shows us in OpenAI's 2020 research paper on AI Scaling Laws and later Deepmind's paper correcting their math, that the diminishing returns will hit a limit far before convergence.

In addition to that, the AI also cannot detect subtle changes to established problems or any new unaccounted for variables, because they're a statistical model and not capable of actual thought. They also lack any sense of responsibility for their actions for the same reason.

You fucking sloppers always try to say "HuMAnS mAkE misTAKeS, TOO!" Yeah and the fucking slopbots are trained on those mistakes and make them again but worse.

[-] FiniteBanjo@programming.dev 10 points 9 hours ago

It's so nice to see some people speaking reason. If only any of those people ran multibillion dollar companies.

[-] FiniteBanjo@programming.dev 1 points 1 day ago

My apologies but after talking with the team about finances the project got shelved. I am sorry if I got anybody's hopes up. I'll be sure to go check my email, now.

[-] FiniteBanjo@programming.dev 12 points 3 days ago

We should be surprised that there are still people at google decent enough to leave.

[-] FiniteBanjo@programming.dev 9 points 3 days ago* (last edited 3 days ago)

I bet something he didn't mention is he probably used a privacy distro on his personal device, and since Google no longer provides the Device Tree in ASOP it's left completely insecure, unable to update.

[-] FiniteBanjo@programming.dev 42 points 3 days ago

Crowdstrike as in the compant responsible for a global outage for machines using it including airports, hotels, fuelstations, banks, broadcasting, and manufacturing?

The company that accidentally made every impacted machine boot-loop because they accidentally added a whole bunch of empty lines of code to production?

47
submitted 5 days ago* (last edited 1 day ago) by FiniteBanjo@programming.dev to c/webdev@programming.dev

My team has decided to put the launch on hold, for now, I appreciate everyone's interest in the project and I'm very thankful for all of the input and helpful discussion.


I tried making a Freelancer post, as I used their service once years ago and it was generally pleasant, but I was immediately bombarded with AI Applicants and an AI Recruiter grilling me for answers to its questions. Immediately deleted the post and the entire account.

Where do I go to get a quote for some work done by some actual human beings? I don't need AI, I don't want AI, if someone automate their quoting process then I'm out because I'm not gonna fuckin negotiate with the hallucinating vending machine.

EDIT: ~~I attempted a post on UpWork.com and there is far less AI response than on Freelancer, but I've still declined all offers so far because they failed an important filter question. Still, I've been able to get quotes! FINALLY! Looks like my website could potentially be built for some number between $5,000 and $38,000. Very interesting. As some commenters have suggested, I'll check out some options in local cities, though I doubt I'll find much in my small studio's price range.~~ EDIT2: These quotes are bogus, I think. I've been seeing the same phrases repeat across multiple proposals "AI is a powerful tool but no replacement for creativity". I went through so many proposals and not a single one looked like a real human capable of adhering to a no-ai policy.

15

I've repaired this worthless older generation Logitech G Hero mouse 5 times in two years, but after its most recent cleaning the 5 pin cable molex snapped and I feel absolutely no desire to keep this creature alive for another moment.

I had a similar cable failure on a Logitech keyboard awhile back, which admittedly I did fix and do plan to keep around because it's hard to find a mechanical keyboard with an aluminium body and also NOT completely covered to the teeth with rainbow LEDs.

Can anybody recommend me a good durable mouse, preferably not aimed at gamers?

18

This already exists for automatic AI detection with a percentage accuracy, but it doesn't usually include a nightshade indicator. There are also different sites that let you mess with contrast and/or saturation, though not all of them work with URL and no site does all three.

The purpose of the contrast/saturation controls are that it makes noise in the image super visible which can be a dead giveaway for AI, but there is also noise in Nightshade.

11
submitted 5 months ago* (last edited 5 months ago) by FiniteBanjo@programming.dev to c/hardware@programming.dev

The PAS 5500/1150C is capable of producing wafers at a resolution of ≤ 90 nm with a wavelength of 193 nm, according to THIS DOCUMENT. It's a machine from the 90s and gets support through 2035.

I don't know what the actual requirements are for printing more modern chips and wafers, though.

Do you think there is much margin to be had with the more recent machines, as in cost vs benefit? There are no import restrictions in my case, for the record.

EDIT:

I did some digging and probably the answer is "NO" because the first 1GB DDR5 from Hynix was ≤ 50 nm and more modern chips use ≤ 20 nm, while I can't find anything confirming lower resolutions can't I doubt any current plans exist for it.

5
submitted 5 months ago* (last edited 5 months ago) by FiniteBanjo@programming.dev to c/webdev@programming.dev

I'm looking at building a website to host comics, a small blog, and store user credentials and comments. Possibly a store.

I've tried this on one separate occasion over a year ago, first I tried using .net as a full stack but I got frustrated with how none of the tutorials on setting up the database, with some forms to submit to it, worked in the then current versions. After that I attempted to program everything in React, but React Router wasn't working well at the time and in general it's more specialized in single page applications. I have hosted some multipage react sites on Ionos before, domains bought elsewhere, so there is no issues on figuring that part out.

So if you were to build it, what would you use? If you were to pay for something like it, what do you think would be a reasonable price?

view more: next ›

FiniteBanjo

0 post score
0 comment score
joined 5 months ago