👁️ Very glad to present our Art/Perception collaboration with Etienne Rey today!

🔗 laurentperrinet.github.io/talk…

What happens today:

We will discuss with Etienne about the emergence of our collaboration, discover the fantastic world of neuroAI, and see and discuss some actual works that emerged from that collaboration

#DeepLearning #ComputerVision #AI #Research #NeuralNetworks #NeuroAI #OpenScience #artScience

A New Look for Convolutional Deep Networks | Next-generation neural computations

By construction, deep convolutional networks treat all pixels equally. Result: they are energy-hungry, fragile to rotations or zooms, and poor at object localization.

^{Next-generation neural computations}

The barrier to entry for powerful AI keeps falling. This demo combines accessible tools (spreadsheet, open models, cloud hosting) to solve a real business problem of products taxonomy classification on scale. The future is modular and affordable. #Accessibility #Tech #AITools #AIFormula #SpreadsheetAI #AI #LLM #WebSearch #SearXNG #PerplexityAI #Ollama #OllamaCloud

matasoft.hr/QTrendControl/inde…

It always amazes me how many people cant be bothered to change their default search engine. It is by far the easiest way to "stick it to the man".

I have been using Duck for the past 10 years. Occasionally I might check with the other engines if I cant find something but has become vary rare. The NOAI version only strengthens the case.

#google #duckduckGo #searchengines #ai

DuckDuckGo is running a poll on whether people are in favour of "AI" or not.

duckduckgo.com/vote/

#AI

Are you YES AI or NO AI? Vote now.

Big Tech doesn't care if you want AI or not. They should.

^{duckduckgo.com}

Thinking about AI costs? This demo used GLM 4.7 model in Ollama Cloud, noted as a 'fraction of the cost' of commercial APIs. Effective AI doesn't have to break the bank. Utilize powerful AI-driven formulas in your spreadsheet, by neglible price.
#CostEffective #AIonABudget #AI #LLM #WebSearch #SearXNG #SpreadsheetAI #Ollama #OllamaCloud #AITools #AIFormula

matasoft.hr/QTrendControl/inde…

This is quite a piece

planetearthandbeyond.co/p/real…

#AI #LLM #salesforce

If anyone ever tries to tell you LLMs are just as good (or better!) in generating text (or code) as humans are in creating text (or code), ask them about "dogfooding".

Dogfooding means training LLMs on their own output. It is absolutely disastrous to such models:
nature.com/articles/s41586-024…

Every "AI" company will have layers upon layers of defenses against LLM-generated text ending up in training data.

Which is why they desperately seek out any and all human-created text out there.

#AI #Hype

AI models collapse when trained on recursively generated data - Nature

 Analysis shows that indiscriminately training generative artificial intelligence on real and generated content, usually done by scraping data from the Internet, can lead to a collapse in the ability of the models to generate diverse high-quality out…

^Nature

Remember how we were all supposed to be "left behind" if we don't jump on the Metaverse bandwagon? Especially businesses?

Yeah, about that:
theverge.com/tech/863209/meta-…

But today we should treat absolutely seriously all the bullshit about "being left behind" if we don't adopt "AI"! 🤡

#Metaverse #AI #Hype

Meta has discontinued its metaverse for work, too

Meta is discontinuing Horizon Workrooms and Meta Horizon managed services, amidst its many VR layoffs.

^{Sean Hollister (The Verge)}

Over €307 million in funding are now available for #AI and related technologies.

Two new calls have just been launched under the Horizon Europe Work Programme, mobilising €307.3M to strengthen Europe’s digital innovation & competitiveness.

Apply now: link.europa.eu/Fftyjc

»curl — Projekt beendet Bug-Bounty-Programm:
curl-Maintainer @bagder hat das Ende des Bug-Bounty-Programms angekündigt. Unbrauchbare KI-Meldungen nahmen wohl überhand.«

Ach was, die KI ist künstlich aber nicht intelligent oder was nun?!?? Ich bin sogar der Meinung, dass dies was die KI angeht noch das rel. kleinste "Problem" ist. Schade dass deswegen das curl Bug-Bounty aufgelöst wird.

🧑‍💻 heise.de/news/curl-Projekt-bee…

#curl #ki #bugbounty #unbrauchbar #ai #uberhang #ausserkontrolle #it #ittools

curl: Projekt beendet Bug-Bounty-Programm

curl-Maintainer Daniel Stenberg hat das Ende des Bug-Bounty-Programms angekündigt. Unbrauchbare KI-Meldungen nahmen wohl überhand.

^{Dirk Knop (heise online)}

At Vivaldi we continue to make choices that are different from our competitors. We have chosen to not integrate AI or crypto, but instead we integrate a wealth of other features, based on the wishes of our users.

We are a European company with most of the team based in Norway and Iceland, a few around Europe and a couple in the US.

Our servers are based in Iceland.

If you want to get away from Big Tech, maybe give us a try? If you are already using Vivaldi, maybe introduce your friends?

Have a nice day!

vivaldi.com

#Vivaldi #Browser #Windows #Macos #Linux #AI #Europa #Technology #EU #UK #Germany #France #Poland #Denmark #Norway #Iceland #Greenland

Vivaldi Browser | Powerful, Personal and Private web browser

It’s a web browser. But fun. It comes with a bunch of clever features built-in. It’s super flexible and does not track you. Get the Vivaldi browser for desktop, mobile, and your car!

^{Vivaldi Technologies}

AI has delayed the planned retirements of at least 15 coal plants, some indefinitely: e360.yale.edu/digest/ai-coal-n…

The head of the 2nd largest utility company in the U.S. says: "We're going to extend coal plants as long as we can because we need those resources on the grid."
#energy #AI #climate

A.I. Is Keeping Aging Coal Plants Online

^{Yale E360}

I noticed that more Gmail spam is getting through, but most are still being flagged. I think their spam classification is only partially driven by Gemini #AI

If they were concerned about spam, they would allow granular application of Smart Features, but we're the product.

#enshittification

It appears that turning off the new “Smart Features” in #GMail that “provide smart features and personalize your experience” also turns off a decades-old feature that categorizes spam away from your Inbox.

Smart Features uses Gemini #AI to enable LLM-assisted email composition and summary, and to suggest stuff from your calendars and promotional items.

In other words, disabling GMail’s AI features requires you to give up a categorization feature that you’ve had for decades, long before “AI” as we know it today was even a thing.

#enshittification

support.google.com/mail/thread…