Brains of birds synch while singing
Links for 28/2023
Non-AI stuff
Pompeii archaeologists discover 'pizza' painting: The flatbread depicted in the 2,000-year-old fresco "may be a distant ancestor of the modern dish", Italy's culture ministry said. It seems to be a circular variation of the Focaccia with toppings. Modern Pizza has been around since the 18th century, and if this round flatbread counts, it would predate modern Pizzas by 1800 years. My favorite food has quite a history!
Related: “The interior angle of the pizza slice emoji in the Apple emoji set 🍕 is 48.5556º, implying that the pizza was cut into a very awkward 7.4 slices.“
Demoscene accepted as UNESCO cultural heritage in The Netherlands — Congrats, and greetings to my old c64-pals from Fairlight, Hotline, Ikari and Genesis*Project.
The Psychedelic Cryptography Contest, in which people hide messages in whibbly whobbly visual stuff that only tripping people high on psychedelics can see.
David Chalmers and Christof Koch settled their 1998 bet if neuroscience can explain consciousness within 25 years. It’s still very much completely unsolved. Also: Thousands of species of animals probably have consciousness.
The Thermonator Is a Robot Dog With a Flamethrower. Buy it here and spread the love.
A white supremacist ceased to be a white supremacist after taking a dose of MDMA in a study on drug use, which is surprising because XTC releases oxytocin — the supposedly “love-hormone” —, which also fuels ethnocentrism and coordinated out-group attacks during conflicts, both of which, well, are a mainstay of white supremacists.
Orcas are further spreading their attack-boats-meme and “rammed a yacht in the North Sea off Shetland“, while a pod of the whales “bumped one of the boats in an endurance sailing race as it approached the Strait of Gibraltar“. The attacks started three years ago and it seems that there is one whale mother spreading that behavior among her offspring. It’s fascinating to watch memetics in animals.
And while we’re at animal memetics: The brains of birds synchronize when they sing duets.
Related: Brain Waves Synchronize when People Interact, which is a well documented but not too well understood phenomenon by now. I wrote about this a couple of times and i suspect that social media enables interbain synchrony via oxytocin, which might be the most simple explanation of the hightened tribalism on socmed.
Related: Scientists discover spiral-shaped signals that organize brain activity: "One key characteristic of these brain spirals is that they often emerge at the boundaries that separate different functional networks in the brain. Through their rotational motion, they effectively coordinate the flow of activity between these networks."
Octopuses, crabs and lobsters to be recognised as sentient beings under UK law. And: Octopuses sleep—and possibly dream—just like humans. Here’s a clip of a sleeping octopus from one study about dreaming mollusc “where Neuropixels recordings shows a wake-like stage of 'active sleep' accompanied by rapidly changing skin patterns that match those of waking behavior“, and that short clip is taken from this neat explainer video from Nature. Also, here’s a thread on octopus consciousness. Also, “Researchers mapped neural activity in an octopus’s visual system, revealing striking similarities to humans.“ Cephalopoda are people too!
Excellent entry to the 36 Days of Type-contest 2023
Chuck “No Pattern“ Anderson still got it: “Exploring frame-by-frame animations using just Illustrator and Photoshop. Complex blend of 100's of shapes along a zig-zagging path –> Envelope Distort (Arch), changing the source path size from big to small and pasting one step at a time into Photoshop as a smart object into a frame animation.“
“In Japan, these grandparents didn't want Kids to feel lonely while waiting for the bus. So, they made a life-sized version of Totoro at the bus stop“.
Mechanical Apple Watch From Real E-Waste Apple Watch. Reminds me of the unsmart rotary mobile phone.
I want T-Shirts featuring illus from vintage japanese ice cream and popsicle bags.
A Ferrofluid drum synth by the unmatched synth furniture designer Love Hulten. Should go well with this Ferrofluid display cell bluetooth speaker.
Lowtech Magazine rebuilt their solar powered website. If i ever do a self hosted webmagazine again, i’ll try to go a similar route.
Anthropocening the shit out of… groundwater: We’ve pumped so much groundwater that we’ve nudged the Earth's spin.
GenZ TikTokers are going anti-nihilist. I see this as a collective response to the teenage mental health crisis caused by social media, and as a necessary countercultural movement to the tired nihilism of the chan-influenced shitposters.
Banksy’s first solo show in Glasgow: Cut & Run. Here’s a walkthrough by Juxtapoz Magazine. There’s also a Banksy-show running in Frankfurt/Main in germany which only shows reproductions of his works.
Trailers for the fantastic looking Landscape With Invisible Hand by Cory Finley, the meme-stock movie aptly titled Dumb Money, Netflix’ adaption of Liu Cixins Three Body Problem, for Poor Things by Yorgos Lanthimos (The Lobster, Dogtooth, The Killing of a Sacred Deer) who blows me away with pretty much everything he does, and to the unlikely-but-real eighth season of Futurama.
AI stuff
The lawsuits against AI-companies keep on coming: ChatGPT maker OpenAI faces a lawsuit over how it used people’s data, Authors sue OpenAI for copyright infringement, Sarah Silverman is suing OpenAI for copyright infringement. Johan Brandstedt has a three part overview about the copyright complaints and it breaks down to: You can infringe copyright for research, but only very limited for commercial products. AI very much are commercial products, thus OpenAI and Stability and Midjourney are in superhot waters. That’s one of a shitcake for MS/OAI and we’re looking into a bright future.
We’ve been here one year ago, when Andy Baio called this behavior ‘data laundering’ and he’s right. The seperate privacy issue is the icing on the cake and the ongoing defamation lawsuits are the cherry on top.
Also, stuff like An AI Is Inventing Fake Quotes by Real People and Publishing Them Online are a million defamation lawsuits in the making. Good luck, Sam Altman, you gonna need it.Study about model collapse for image synthesis: Self-Consuming Generative Models Go MAD. Image models trained on their own data show diminishing returns, which is mainly a compression loop in which a JPGd image detoriates after each compression. There are some good examples in the paper, but i think you can see the detoriation most clearly with the typographic example:
Douglas "Gödel, Escher, Bach" Hofstadter recently appeared on a podcast voicing concerns about AI, here’s some excerpts at LessWrong. In The Atlantic, he now wrote down those concerns: Gödel, Escher, Bach, and AI: "I frankly am baffled by the allure, for so many unquestionably insightful people (including many friends of mine), of letting opaque computational systems perform intellectual tasks for them. (…) To fall for the illusion that vast computational systems “who” have never had a single experience in the real world outside of text are nevertheless perfectly reliable authorities about the world at large is a deep mistake, and, if that mistake is repeated sufficiently often and comes to be widely accepted, it will undermine the very nature of truth on which our society—and I mean all of human society—is based."
Bruce Sterling on modern AI-mythology: "these beasts, the Basilisk, the Masked Shoggoth, that Paperclip gizmo, they were born from a trillion dimensions. No wonder they impress. Some critics call them mere parrots built with fancy mathematics: ‘stochastic parrots.’ A Large Language Model is built from complex statistics, so it's a parrot yakking up its slurry of half-stolen words and images. But those ‘parrots’ are also AI mythic beasts—parrots with a trillion dimensions. It's as if that ‘dead parrot’ in the legendary Monty Python sketch could take your job, or burst right out of the BBC-TV screen like a blazing phoenix and eat the television signal. Those parrots are dynamite!"
"Looks like it's back to his roots, on the forthcoming Aphex Twin album ‘Kick out the Jams, Mother Folkers’". Now I wonder what a folk cover of Come To Daddy would sound like.
Another paper about generating AI-images from brainwaves: DreamDiffusion: Generating High-Quality Images from Brain EEG Signals. Tech like this has been making the rounds in the last months and i wrote about this stuff in context of Neuralinks approval for human BCI-trials some weeks ago.
Evgeny Morozov in the NYTimes: The True Threat of Artificial Intelligence is hegemonial neoliberalism. I’m not sure if there is one "true threat", but he’s also right that corporations sucking up the knowledge of the world for free and reselling a stochastic version to us for a buck is a threat.
An Indian politician says scandalous audio clips are AI deepfakes. We had them tested: "Deepfake experts noted that AI can be used as a cover by politicians when embarrassing audio and video clips of them emerge."
I wrote about the dangers of self radicalization through chatbots before and now the Guy Who Tried to Kill the Queen of England Was Encouraged by AI Chatbot.
Archaeologists Use Artificial Intelligence (Ai) To Translate 5,000-Year-Old Cuneiform Tablets: "A team of archaeologists and computer scientists have created an AI program that can translate ancient cuneiform tablets instantly using neural machine learning translations."
Me, in January: "What AI is automatizing here is George Carlins ‘big bullshit businessmen smile’". Those smiling businessmen a week ago: Linkedins new AI will write hustle bro posts for you. ¯\_(ツ)_/¯
Dildo makers are getting into AI: "The ‘Advanced Lovense ChatGPT Pleasure Companion’ invites you to indulge in juicy and erotic stories that the Companion creates based on your selected topic. Lovers of spicy fan fiction never had it this good, is all I’m saying. Once you’ve picked your topics, the Companion will even voice the story and control your Lovense toy while reading it to you." Give it to me, ChatGPT!
"Using Photoshop’s generative fill to extend Star Wars so it works vertically".
AI-Spam is going strong: Junk websites filled with AI-generated text are pulling in money from programmatic ads and AI-Generated Books of Nonsense Are All Over Amazon's Bestseller Lists: "This will absolutely be the death knell for [Kindle Unlimited] if Amazon cannot kill this off".
The Vatican Releases Its Own AI Ethics Handbook, download the thing here. I haven’t read it, but I wonder if it has sections on the second commandment "Thou shalt not make unto thee any graven image" and/or about "Thou shalt not make unto thee stupid images of the pope in ridiculous white puffer jackets".
"I got banned from Midjourney AI for generating realistic images of politicians cheating on their wives for a series called ‘AI will revolutionize the blackmail industry’". And AI just ate the job of Alison Jackson.
"A group of San Franciscans realized that they can disable Waymo and Cruise robotaxis by placing a traffic cone on the vehicle's hood. They're now encouraging others to do it: 'Hell no, we do not consent to this.'"
Cool image synthesis by Airminded AI about parapsychological experiments from the soviet era.
On Twitch, You Can Watch AI Trump Vs. AI Biden In An Endless, Unhinged Live Debate. It goes exactly how you think it goes: "Both AI versions of the politicians curse heavily at each other: at one point I heard Biden call Trump a limp dick and Trump retorted by telling him to go back to jacking off to Charlie and the Chocolate Factory."
Generative AI coming to Wordpress Jetpack. Reminder: As of 2022, a staggering 43% of all websites on the web run on Wordpress and Jetpack is their monolith plugin.
In a sponsored video, the popular Machine Learning Street Talk-podcast talked to YouAi founder Dmitri Shapiro who "believes digitizing our inner lives could unlock tremendous benefits" and provides an AI-platform which learns from user behavior to "develop a multidimensional model of their mind". It’s exactly this stuff that gives me the heebie-jeebies. I have a very bad feeling about AI-twins of real people. While i don’t think there should be jailtime for corporations that train counterfeit people, like Yuval Noah Harrari suggested while talking at UN’s AI for Good global summit in Geneva, i do think that exactly this stuff has the potential to have some real grave consequences for the human understanding of themselves.
Study: AI model GPT-3 (dis)informs us better than humans: "In comparison with humans, it can produce accurate information that is easier to understand, but it can also produce more compelling disinformation."
I’m beating the drum against Open Sourcing AI for a while for various reasons, and now researchers looked into the security vulnerabilities of the "50 most popular Large Language Model (LLM)-based projects on GitHub". Not surprisingly to me, they found that "None of the projects analyzed scored higher than 6.1 – the average score was 4.6 out of 10 – indicating a high level of security risk associated with these projects, Rezilion said. In fact, Auto-GPT, the most popular project (with almost 140,000 stars), is less than 3 months old and has the third-lowest score of 3.7, making it an extremely risky project from a security perspective."
Infinite Photorealistic Worlds using Procedural Generation shows where stuff like No Man's Sky is heading if you add modern computer vision stuff into the mix.
Here come the “metahumans”: Virtual avatars have real jobs in Indonesia: "Metahumans are occupying trusted roles like news anchors and government communicators."
CloneDub — All your podcasts will be available in any language soon, in the voices of your podcast hosts.
How the researchers behind the paper MotionGPT: Human Motion
as Foreign Language didn't exemplify this research at least with a left leg doing a forward aerial half turn every alternate step is beyond me.
Thank you. Thank you. Thank you. 💜