₦฿₭
banner
nbk.bsky.social
₦฿₭
@nbk.bsky.social
Paw / Pwn / Purr
@nbk_2000
I think few people who understand how LLMs work believe they're capable of consciousness

For that matter, i doubt anyone after even reading Stephen Wolfram's article on ChatGPT could be persuaded that this algorithm "deserves to vote!"

Regardless, LLMs will let us know when they're conscious 😋
March 3, 2025 at 4:59 AM
Seems this noodling is in the zeitgeist

youtube.com/watch?v=biUf...
Gödel's theorem debunks the most important AI myth. AI will not be conscious | Roger Penrose (Nobel)
YouTube video by This Is World
m.youtube.com
March 3, 2025 at 3:40 AM
If they'd stop addressing me by my full name and accurately describing what i dreamed the night before, that'd help change my mind. Until then, i guess they're omniscient 🤷‍♂️

What makes you doubt the great algorithms?
February 18, 2025 at 12:35 PM
Making TI go brrrrr 🐾
February 13, 2025 at 2:11 PM
That's pretty sick! This would be especially powerful when combined with confidential data i also need to analyze. Have you looked into who gets access to your source material?
February 10, 2025 at 3:22 AM
That content resembles the reason people give for leaving Twitter... perhaps the platform isn't the problem😐
February 7, 2025 at 1:09 AM
Haha, I read your isolated comment 3 times trying to figure out what terrible things befell poor Guido van Rossum that I'd been unaware of 🤣
January 7, 2025 at 1:59 PM
I'm guessing... Bob's his uncle 😹

... apologies, couldn't resist.
January 4, 2025 at 1:42 PM
Wait a minute, that's a piña colada!!! 😹
January 2, 2025 at 2:00 AM
Go drunk, you're home 😹
January 2, 2025 at 1:51 AM
This can be solved by giving the stats of the reviewer. If Newb1 with 100 N/As says a company sucks, hackers are smart enough to read between the lines.

It's a problem BEST solved by the platforms but this has been requested for YEARS so apparently its not in their interests to protect researchers.
December 28, 2024 at 6:26 PM
Do we only solve easy problems these days? 😋

That said you wouldn't need to monitor anything. The agreement about sharing private info is between the hunter and the program. It's on the hunter if they don't have permission to tell others that the program is awesome... or abusive.
December 28, 2024 at 6:18 PM
A rating system that hackers can use to provide public feedback about programs.

Performing speculative work for companies that may, or may not, abuse you feels gross.
December 28, 2024 at 4:08 PM
Mine is literally NaN... seems about right 😹
December 28, 2024 at 3:59 PM
Hoping all those years of manually brute forcing kiosks will pay off?
December 23, 2024 at 11:21 AM
Fixed it! "Artificial Intelligence Detection Service"
December 23, 2024 at 11:15 AM
I'd like to talk to you about Firefox Hoodies, will that work? 😸
December 17, 2024 at 2:37 PM
Real world answer from a former CISO? Number 1. Because pentests are for compliance and might be shared with clients. Investing in the Red Team and BB program for actionable findings provides better ROI.
December 9, 2024 at 1:29 PM
👆Yep, this.

Now that you know what they are, you can develop the habit of skipping them, safely knowing that you'll read the content later in the article.

This'll also help you skip links-disguised-as-subheadings, which have a similar visual pattern (and a typographical affront worthy of a rant)😋
December 9, 2024 at 1:29 AM
Pruning is key to keeping a high quality signal:noise ratio on your feed. Using techniques like this, even my Brand X feed was decent. BSky also has a global "show reposts" filter. Once you follow enough original content creators, you can skip noise that comes from people mindlessly tapping repost.
December 9, 2024 at 1:10 AM
Well, barring a major upset, it appears the Twit faction has outvoted the bsky this round. So we'll be doing the AMA over there... this time!

Check @nbk_2000 on brand X for the details 🐾
December 3, 2024 at 8:59 PM