After watching OpenAI's presentation and, even more, an excerpt from Google's presentation yesterday, I asked myself: is this AI, according to the big tech companies (especially Google - for OpenAI it's their core business, so I understand their perspective), truly what users want and need, or is it just another method to lock people into using their technologies, which are not easily self-hostable?
I'm not arguing for or against it, but I noticed that (almost) the entire Google I/O yesterday was focused on this...
#AI #BigTech #GoogleIO #OpenAI #Technology #Innovation #UserNeeds #SelfHosting #TechIndustry #FutureTech
Sh*tf*ck 😳🔥😔
#ChatGPT consumes 25 times more #energy than #Google
"Energy consumption by Artificial Intelligence (#AI) is rising rapidly: AI is predicted to consume twice as much energy as the whole of France by 2030, according to some calculations."
https://www.brusselstimes.com/world-all-news/1042696/chatgpt-consumes-25-times-more-energy-than-google
InstructLab Demo: Lowering the barrier to AI model development
The June 2024 issue has been released. This month, Joe Casad weighs in on #Google's new algorithm and what it's doing to search https://www.linux-magazine.com/Issues/2024/283/Welcome #SEO #AI #algorithm #content #search #bookmark
I think of this EVERY time someone suggests using #AI for anything more than programming boilerplate, which you can copy and paste.
Charles Babbage, inventor of the first mechanical computer, has a particular quote that's always stuck with me:
"On two occasions I have been asked, 'Pray, Mr. Babbage, if you put into the machine wrong figures, will the right answers come out?' I am not able rightly to apprehend the kind of confusion of ideas that could provoke such a question."
Babbage made this observation almost 200 years ago, and he'd certainly say the same thing about modern AI proponents who expect that feeding all of the internet into an over-glorified autocomplete will produce factual results.
I have 3 Easier Ways to Avoid Dissatisfied Customers
1. Disconnect your support telephone number
2. Fire all your call center staff
3. Replace them all with an #AI powered contact form
You’ll never hear from dissatisfied customers ever again.
https://middleme.net/2024/05/10/3-ways-to-avoid-dissatisfied-customers/
This is anecdotal, but I now personally know two employees of #AI-related companies whose valuations have imploded within the past month, and they are now looking to bail. One company lost 73% of their (publicly traded) value in a month. The other one is a startup whose founder (according to my friend) has “finally snapped”, and the company is now in freefall.
It feels like the bubble is bursting right now.
Introducing Podman AI Lab: Developer tooling for working with LLMs
https://developers.redhat.com/articles/2024/05/07/introducing-podman-ai-lab
FFS!
Stack Overflow bans users en masse for rebelling against OpenAI partnership — users banned for deleting answers to prevent them being used to train ChatGPT
A new Study from Tesearchers at MIT and the University of Washington reveals an AI Model that can accurately predict a Person or a Machine’s future Actions #Research #AI
- Paper on OpenReview https://openreview.net/forum?id=W3VsHuga3j
- Article https://interestingengineering.com/innovation/mit-ai-predict-human-actions
“So AI is a lying machine made out of crimes that’s destroying the earth and costs a fortune and is being deployed recklessly everywhere without any consideration to your safety. But is it a scam? Yep.”
@ben #Funfact: all "#learning" is #FairUse, otherwise you'd be a perpetual #DebtPeon to ]whoever made your schoolbooks and created whatever media you ever consumed](
http://felixreda.eu/2021/07/github-copilot-is-not-infringing-your-copyright/ ) !
- Notwidthstanding the fact that #WastefulComputing for wannabe "#AI" is just marginally less bad than #Shitcoin-#Cryptocurrencies...
-->AI
"A few years back, a writer in a developing country started doing contract work for a company called AdVon Commerce, getting a few pennies per word to write online product reviews.
But the writer — who like other AdVon sources interviewed for this story spoke on condition of anonymity — recalls that the gig's responsibilities soon shifted. Instead of writing, they were now tasked with polishing drafts generated using an AI system the company was developing, internally dubbed MEL."
Come check out InstructLab AI, an open source, open community, open contribution development workflow for training LLMs! It's based on the new LAB training strategy and can be combined with things like RAG. All powered by open source and truly open licensed models.
We get a lot of AI trash. But sometimes some Internet citizens get annoyed by a thing and now they can make songs about it with suno AI. And they give their anger about Corporate BS quite some energy and fire.
Enjoy the Song
Today llamafile (project by Mozilla’s innovation group) is both the easiest and fastest Way to run a wide Range of open Large Language Models on your own Hardware - Blog post by Stephen Hood @stlhood, Mozilla open source AI lead #LLM #AI https://hacks.mozilla.org/2024/04/llamafiles-progress-four-months-in/
Benchmarks of 14 LLMs in an Chatbot Arena with Street Fighter III video-game - Project by Banjo Obayomi #AI #LLM https://community.aws/content/2dbNlQiqKvUtTBV15mHBqivckmo/14-llms-fought-314-street-fighter-matches-here-s-who-won
Of course #Microsoft needed a fusion-powered #AI.
How else do you think they planned to push their new subscription model and the Start Menu's adverts? /s
If you have a website with neo cities, just know that the creator actually likes AI, and will add it again based on this post alone. Moratorium on freaking out about AI https://kyledrake.com/writings/ai #AI #IndieWeb #SmallWeb #NeoCities
The more I look around, the more I realize that many people making important decisions haven't grasped what Large Language Models (LLMs) are. Recently, a potential client asked me to set up a server with absurd requirements (using Hyper-V to host Linux VMs inside a Windows VM). When I requested explanations, they told me this solution had to be valid because it was "suggested by AI".
When I boil it down, I find my feelings about AI are actually pretty similar to my feelings about blockchains.
#ArtificialIntelligence #AI #newsletter #CitationNeeded
I spent a long time experimenting with AI before finally writing about it in depth. It can be pretty useful — but is it worth it?
#UK competition #watchdog has 'real concerns' over big tech #AI dominance
‚The CMA found an "interconnected web" of AI partnerships involving the same firms: #Google, #Apple, #Microsoft, #Meta, #Amazon, and chip-maker #Nvidia.
"[…] Now, with a deeper understanding and having watched developments very closely, we have real concerns“‘
Gartner: 75% of enterprise software devs will use AI in 2028 • The Register
„While the popularity of #AI #coding tools will undoubtedly increase, development teams might need to manage the expectations of their senior managers ... 'They're not going to hear it from the vendors'“
https://www.theregister.com/2024/04/13/gartner_ai_enterprise_code/
Degitalisierung: Infinite Money Glitch
Voll innovativ..
„Weil das alles aber nicht so schnell Energie bringen wird, freut sich die Fossilindustrie über die steigende Nachfrage nach Gaskraftwerken […].
Der Weg in die vermeintliche Zukunft geht […] über das energietechnische Mittelalter.“
#AI #KI #energieverbrauch #LLM #aihype
https://netzpolitik.org/2024/degitalisierung-infinite-money-glitch/
How's that #AI Right to be Forgotten coming?
#FollowTheMoney 🧵 16/n Back to who the money flows towards:
Nobody had a better year than Meta CEO Mark Zuckerberg, whose net worth has increased by more than any other billionaire, up by nearly $113 billion over the past 12 months, to an estimated $177 billion #AI #Meta
Python OpenAI demos: a collection of Python scripts that demonstrate how to use the OpenAI API to generate Chat Completions - Project by Pamela Fox @pamelafox #AI #LLM https://github.com/pamelafox/python-openai-demos
#AI: Pollute our data, pollute our environment.
Interesting discovery: ever since I swapped out the Samsung Keyboard for Gboard, my Samsung S24 Ultra feels noticeably sleeker. Starting to think all those 'AI Features' might actually bog down the device rather than enhance it.
https://www.972mag.com/lavender-ai-israeli-army-gaza/
Skynet is now online and slaughtering people in Gaza with gleeful abandon. Humans in the IDF only rubber stamp the AI’s decisions or act as meat puppets to drop #AI mandated bombs on children.
Ah well. Gotta work out the kinks before these toys are deployed in the imperial core, after all!
LOL! So a few blindness higher ups want me to speak at their library events about the use of #AI and how it could enhance the lives of Disabled people, so I go through the song and dance, arranging payment, not at all hiding the fact I am very critical of AI in the disability space because I assume they read my blog by now, but then they come back and say things like, you didn't tell the internet you are actually against AI, to which I send them 5 blog posts where I critically talk about AI and they are always like but we didn't know this!
Pro tip for others. Read my blog, not just social media. That's what my website is for, after all.
LangChain Go: the Go language implementation of LangChain to build Applications with LLMs through Composability #AI #Programming https://github.com/tmc/langchaingo/
#YossiSariel unmasked as head of Unit 8200 and architect of #AI strategy after book written under pen name reveals his Google account
The identity of the commander of #Israel’s #Unit8200 is a closely guarded secret. He occupies one of the most sensitive roles in the military, leading one of the world’s most powerful surveillance agencies, comparable to the US #NSA https://www.theguardian.com/world/2024/apr/05/top-israeli-spy-chief-exposes-his-true-identity-in-online-security-lapse
Tech stocks go ever higher, but how much of that is due to real gains, and how much is because of layoffs done in the foolish hope that generative #AI tools can replace workers?
https://www.computerworld.com/article/2077751/is-ai-driving-tech-layoffs.html by @sjvn
DarkGPT: an OSINT assistant based on GPT-4-200K designed to perform queries on leaked Databases, thus providing an #AI assistant that can be useful in your traditional #OSINT processes https://github.com/luijait/DarkGPT
Google DeepMind’s TacticAI can predict the Outcome of Corner Kicks and provide realistic and accurate Tactical Suggestions in Football Matches #Sports #AI https://www.technologyreview.com/2024/03/19/1089927/google-deepminds-new-ai-assistant-helps-elite-soccer-coaches-get-even-better/
What will be included in the #AI Olympics?
Making me laugh is a good start.
Can your #AI put together a stand-up routine that brings joy and laughter?
Dear #artists #artist #designers #designer #graphicdesigners #graphicdesigner #illustrators #illustrator #painters #painter
I found these two tools that should help protect your work from #abusive #AI scraping and training; These tools are called #Glaze and #Nightshade both can be used simultaneously!
The former protects your images, the latter poisons the (abusive) models.
Very useful and important to know and to share:
https://glaze.cs.uchicago.edu/what-is-glaze.html
https://nightshade.cs.uchicago.edu/whatis.html?ref=journal.hexmos.com
A big thank to @informapirata that made discover these essential tools!
My opinion? Yes. We have reached peak #ai. https://www.wheresyoured.at/peakai/
It has been a month that I am using daily 9to5 a generative #LLM #AI tool and my verdict is...
<h1>what a huge amount of crap it is!</h1>Yes I uploaded more than 1000 products in a month, but the quality it is just ridiculous.
Pretty soon I realized the training material the platform is using, is very limited, and the description tends to be all identical.
These LLM are just sophisticated plagiarism tool and nothing else...
I have long argued that describing LLMs fabricating information as "hallucinations" is a category error—a hallucination being a mistaken perception, and an LLM not being a being with a mind that perceives.
But I'm coming to recognize that I also dislike the term even as a metaphor. A hallucination in a human being is a breakdown in perception caused by extreme circumstances or neurological issues; a hallucination in an LLM is the result of the normal operations of the LLM.
#LLMs #AI
@flyingsaceur alright if you *just* read the first section this was my crude first swing at a world of shouting in lisp in your honour, if you'll have it: #technocognitive #ai #gopher #phloggersGarage #worldOfShouting #lisp #commonLisp #emacs #orgmode
gopher://tilde.club/0/~screwtape/technocognitive/world-of-shouting-1.org
gopher://tilde.club/1/~screwtape/technocognitive/
https://gopher.tildeverse.org/tilde.club/0/~screwtape/technocognitive/world-of-shouting-1.org
https://gopher.tildeverse.org/tilde.club/1/~screwtape/technocognitive/
Elon Musk starts to earn respect from me.
He helped Gina Carano with disney, and now he will open sources for emm... grok?
IDK, maybe i will jump to X? Elon is based. Everyone have cons, but i like him.
btw. if he open 'grok' i wonder how many haters will use it.
My gracious host Doug Ortiz allowed me to babble on his #podcast Tech Bits about my book #PostgreSQL Mistakes and How to Avoid Them, the IT industry and community, society in general, and other light topics. Catch it below:
https://youtu.be/FJMxbt00IJM?si=-7TmETZ_PgDMMBsr
#Postgres #OpenSource #Database #AI #Databases #ManningBooks
Well it seems that Docusign, the largest online signing service for mostly confidential documents... is using your documents to train it's AI regurgitation machines:
https://mastodon.social/@gvwilson/112012277852906749
Thankfully it seems that there's also an open source, self hostable alternative in Documenso:
Https://Documenso.com
Please boost and talk to your org about switching today! (I'm not affiliated)
(But not for me).
But yeah, #BSD based systems like #FreeBSD or #NetBSD still work the same way like like 20 years before. For #Linux - I think #comtainers / #Kubernetes are/were the hot sh…. Things are more moving to #AI I guess… let’s see what the next big thing will be in a few month after the hype.
LinkedIn was down. A lot of people were panicking.
But rather than panic, I saw an #opportunity.
Using all of my strength I ran to the nearest LinkedIn datacenter. I was able to gain access because I made a #personal #connection with the security guard. I actually invested in their ceramics business while I was talking to them.
Once I’d gained access to the servers I was able to deploy a fix I’d written using ChatGPT #AI #genAI.
I fixed LinkedIn, and walked out of the datacenter where everyone was applauding.
I say this not to brag or show off, but to share a story of how you have to show #leadership in the moment, and step up when you can. The CEO of LinkedIn called me that night to thank me. #influencer #hustle #horseownership
"Welcome to SpinScore - an advanced AI tool designed to analyze and score potential biases, logical fallacies, and misleading information in content."
🤔
https://spinscore.io/
Researchers create #Ai worms that can spread from one system to another
I wonder how world with #AI will look like.
Maybe we need dark times, with soulless products around us.
Imagine a situation where your kid came back from school, and say "Hey! look i've drawn a picture for you!"
... you take a look at this handmade picture and realize how priceless it is.
Maybe humans will start to appreciate small things more?
How the mighty have fallen.
I remember c|net as a standard-bearer of tech journalism back in the 1990s. Theirs was one of the few bookmarks I opened regularly to read about tech news.
Red Ventures has destroyed its reputation, along with all their other properties’ like ZDNet (the other fantastic tech news site). RV’s decisive pivot to AI is likely fatal; they will accelerate the descent of all their outlets into a content swill mill fit only for your blocklist.
#AI is an #enshittification accelerator.
“Wikipedia No Longer Considers CNET a "Generally Reliable" Source After AI Scandal”
I don't mind some reggaeton, but only when it's on our terms at home. Seems like a positive use of #AI to me! 
Are you ready to back up your AI chatbot's promises? You'd better be https://theregister.com/2024/02/23/opinion_column/ by @sjvn
If you rely on #AI for work, remember that you're legally responsible for its mistakes. And, boy, will there be mistakes.
«EdgenAI – GenAI made Local: Run Generative AI locally, no internet, completely data-private»
I am not a friend of artificial intelligence, but it may happen in the future that I "have to" use it. Do any of you know GenAI, which is developed in Rust, and if so, how is it?
⚡ https://edgen.co
—
#openai #rust #rustlang #nointernet #ai #OpenGPD #llm #opensource #genai #opensource #oss #privacy #data #LLMs #dataprivacy #coding
While Sam Altman admits that his technology is only possible with stolen intellectual property, implying that his strategy is to steal quickly, and make more money than the lawsuits cost him... The popularizers defend the inexorability of artificial intelligence, because the opposite is democracy and legal system.
I'm not so familiar with #AI and security on the #Internet that I really have a #clue. That's another reason why articles like this are very good for informing me and explaining it to customers. #WebSec is more than just #password and #secrecy.
«Post Quantum Cryptography: Why We Need Resistant Encryption NOW.
Quantum-resistant or post-quantum #cryptography is our best bet against attacks from upcoming quantum computers to increase #security and #privacy.»
– @Tutanota