Artificial intelligence still has some way to go

It feels like the whole internet is reaching a point with AI bots similar to the relationship between Microsoft operating systems and viruses, telephones and robocalls, or Facebook and fascist memes.

its probably pretty hard fwiw because theres a lot of effort being put in to stopping it xp

obvs not impossible tho, some estimates half of all ad revenue is fraud lol

the challenge isn’t making a bot, it’s running them in a way to avoid ban filters. you see those pictures of spam/comment factories where there are hundreds of burner phones stuck to a wall, all being coordinated by a script

yup

i bet if one were enterprising and really got to know the marketplaces for the various pieces necessary to put an ad fraud business together they could prob do it without too much technical ability, then of course you would be committing a felony, maybe ok if you live in russia

maybe in the future you will be able to ask an ai "please do fraud for me" and the police will come directly to your home

it's the ads that are the real fraud if you ask me!

much to think about

you see those pictures of spam/comment factories where there are hundreds of burner phones stuck to a wall, all being coordinated by a script

― ɥɯ ︵ (°□°) (mh), Monday, December 4, 2023 2:11 PM (twenty-six minutes ago) bookmarkflaglink

wonder how they spoof the ip addresses, global botnet?

the period right before the exclamation points actually makes me think this was written by a human

― silverfish, Monday, December 4, 2023 12:57 PM (forty-seven minutes ago) bookmarkflaglink

yeah I thought that too though who knows if it's trained on YouTube comments then the AI might start doing it too. that said I guess the sort of person who would actually watch and comment on a video like this must be pretty weird to begin with

^^ very otm lol

I wonder what it will be like when the audience ofr all internet things is comprised of bots and ai things - will we still love all the likes and attention?

honestly i think people will continue to make a version of the fundamental attribution error - we, as individuals, will tend to think that we know more than average, believing that our own misfortunes are caused by circumstances outside of our control, while attributing the misfortunes of other people to some sort of flaw in their ability to detect bullshit

i of course am on the holy level of fundamental attribution error where i stand in the clouds and shoot lasers at people with my fingertips, pew pew

rude

https://i.imgur.com/5Sx3XCC.png

https://www.bing.com/images/create/audience-that-consists-of-robots----impressionist-/1-656f7cd5359f4c46a4d2b42a23f9435b?id=kL5vdC%2FshbucKTiOnjAiWw.IlKg6ZUePAUwePrABqy%2B3Q&view=detailv2&idpp=genimg&form=GCRIDP&ajaxhist=0&ajaxserp=0

https://i.ebayimg.com/images/g/j5QAAOSw-FRb5y-Q/s-l1600.jpg

Another one for the "these guys are lame" file

Forbes has learned that Guillaume Verdon, the founder of stealth AI startup Extropic and a former Google engineer, is behind the provocative Twitter account leading the “effective accelerationism” movement sweeping Silicon Valley.

Andreessen Horowitz cofounder Marc Andreessen says @BasedBeffJezos is a “patron saint of techno-optimism.” Garry Tan, who cofounded the venture firm Initialized Capital before becoming CEO of Y Combinator, calls him “brother.” Sam Altman, who founded OpenAI — the company that finally mainstreamed artificial intelligence — has jokingly sparred with him on Twitter. Elon Musk says his memes are “🔥🔥🔥.”

https://www.forbes.com/sites/emilybaker-white/2023/12/01/who-is-basedbeffjezos-the-leader-of-effective-accelerationism-eacc

woims as experiences and opportunities

https://i.imgur.com/vegOtUB.png
https://cdn.bsky.app/img/feed_fullsize/plain/did:plc:u7x7lujm5mxy25eeepqgu3nk/bafkreiepmta7p7waomibr34kzlekrxfwzmu3lqcho4wl6sdnq3gfzzjl44@jpeg
https://cdn.bsky.app/img/feed_fullsize/plain/did:plc:u7x7lujm5mxy25eeepqgu3nk/bafkreidodc267tccaem5vrkfcz2u423qs7jzw7mqjxm3bsrmsqdeeapgoe@jpeg
https://cdn.bsky.app/img/feed_fullsize/plain/did:plc:u7x7lujm5mxy25eeepqgu3nk/bafkreieydchggwn3mpbou2khu2qhmgmxdg4cn6uv5snngh4ietqr2vadpu@jpeg
https://cdn.bsky.app/img/feed_fullsize/plain/did:plc:u7x7lujm5mxy25eeepqgu3nk/bafkreigpnode37eihap5m3qutbicnwgl4dkw3bzkzbk7pzu5a23g6yrdxm@jpeg

Samantha Lee right on top of it.

excuse me, i'm looking for a hidden hiking trail, known only to locals. but i only have this jar of woims. will someone trade me a guided tour for the woims?

yes

lmao

That is incredible

https://www.ownyourownfuture.com/they-pay-me-in-woims/

Samantha Lee
Samantha Lee is the founder and editor-in-chief of the Own Your Own Future blog. As a serial entrepreneur and business coach, Samantha is passionate about empowering people to build businesses and careers they love.

After struggling to find fulfilling work in the corporate world, Samantha started her first company at age 30. Through many ups and downs, she gained first-hand experience in creating a purpose-driven business aligned with her values.

Eager to help others find professional happiness on their own terms, Samantha launched the Own Your Own Future blog. She provides tips and interviews with entrepreneurs across industries to inspire readers to take the leap.

With a background in marketing and over a decade running her own companies, Samantha offers practical guidance on topics like getting started, managing finances, marketing effectively, and maintaining work-life balance. She also shares insights from her own entrepreneurial journey.

At the core of Samantha's message is her belief that each of us has a unique set of gifts and dreams. By taking control of our careers and finances, we can spend our time and talents in meaningful ways that serve both ourselves and our communities. Samantha currently lives in Fremont with her family, where she runs her latest startup while continuing to empower budding entrepreneurs.

samantha's website also has a parent's guide. i think i will use this as a resource for raising my children.

https://www.ownyourownfuture.com/high-school-parents-guide/

Samantha's depth and range of knowledge has earned her many plaudits, including the personal endorsement of US Congressional Representative George Santos (R-NY).

samantha is so inspiring

https://i.imgur.com/udGNZFr.png

Ah, Chili’s – the land of sizzling fajitas, bottomless margaritas, and…payday mysteries? Whether you’re a seasoned server slinging plates or a curious cook prepping for their first shift, the question remains: does Chili’s really pay weekly? Fear not, fellow Chili’s comrades!

https://i.imgur.com/PafVDgD.png

they pay in woims unfortunately. i worked there a few summers ago.

leave a big handful of gross wriggly worms as a tip at chilis the servers love it

https://i.imgur.com/5uIgh9P.png

https://i.imgur.com/pFdGOeC.png

rip david shwimmer

Paid them in hideous fingers.

that's how Oliver Hardy says worms - I like that accent. "Woik hard"

i thought this article was interesting, regarding LLMs solving grade-school level math problems they haven't seen before:

https://arstechnica.com/ai/2023/12/the-real-research-behind-the-wild-rumors-about-openais-q-project/

I think I read somewhere that llm can also develop abilities that were not part pf their original training - really I think they will soon enough be able to do do most things a human can do as far as intelligence tasks - after all technology tends to improve in time

you know what they say about assumptions

the idea that any one technology will just continue to rapidly improve forever is IMHO pretty dubious, the typical pattern is rapid initial improvement then slowing incremental progress, airplanes not being that much better than they were 50 years ago is a classic example, of course with ai it doesnt do any human intelligence tasks so we havent even reached the threshold where rapid improvement can begin

i like lagoon's view on this. we won't all be paid in woims just yet!

xp- do you not find copilot useful for coding? still gotta debug it every few times but i've been getting great stuff out of it

yeah i like copilot for single line code completion, even then you gotta keep an eye on it it will do weird things it doesnt care, tho it can be good for longer things if you just dont know the syntax youre supposed to be using, get it to spit something out then try to figure out what its talking about

c0ding ais are interesting, people are like its gonna take the jobs but code is for humans its literally writing something for humans to read, if it really wanted to take the jobs it would just produce a binary, of course some black box binary wouldnt be of much use unless somehow the ai got it perfectly right, and it wouldnt get it perfectly right cause even if it got really good normal conversational speech isnt precise enough instruction to specify a computer program, you need something better, something more purposes built for the job, you need... code, lol, thats my conference speech on why ai will not take the programming jobs, that and also theres fundamentally an infinite demand for programming

there was a thinky lecture upthread i think from some programming dude whose whole point was this, that his view is that LLMs will eliminate the need for programming altogether. you will tell the model what you want and it will do it and nobody needs to look at code. of course like a whole lotta this stuff..... that's a ways off

even then it would still need to talk to a person to get instructions, that person would be a programmer

its pretty funny to think about how much more frustrating the already very frustrating task of debugging would be if you had to talk to the computer that created the bug in order to debug it

however, ai debugging tools, pattern recognition and so forth, in a normal programming workflow will prob be useful

the idea that any one technology will just continue to rapidly improve forever is IMHO pretty dubious, the typical pattern is rapid initial improvement then slowing incremental progress, airplanes not being that much better than they were 50 years ago is a classic example, of course with ai it doesnt do any human intelligence tasks so we havent even reached the threshold where rapid improvement can begin

Yep. There’s no reason to believe that transformers will dramatically accelerate beyond their current state. I won’t discount the usefulness of transformer applications for some problems but the idea LLMs will replace programmers is absurd.