Policy Making in the Post-Truth World - On the Limits of Science and the Rise of Inappropriate Expertise - Rayner and Sarewitz’s warning that we avoid turning models designed to inform problem management into the object of management themselves.
tag > ML
-
-
Eternal Change for No Energy: A Time Crystal Finally Made Real
Google's quantum computing team in collaboration with physicists say they have demonstrated a genuine time crystal for the first time. Like a perpetual motion machine, a time crystal forever cycles between states without consuming energy. Physicists claim to have built this new phase of matter inside a quantum computer.
-
Today in "the fall of western civilization": The 2021 "Global Emerging Technology Summit" of the "The National Security Commission on Artificial Intelligence" (NSCA) (8h Video)
#Comment: This gathering of western military, business and political "leaders" (50%+ spooks) is in essence a total shit-show of the very worst kind. "Diverse" participants - from sleazy used car salesmen, loony dr.strangeloves, to soulless think-tankers and political puppets - are discussing things like A.I, info-warfare and killer robot efforts, all in the name of preserving the "rules based international order" and "democracy" (what ever that means). Fantastic comedy!
-
The Honest Weather Prediction Service
An engraving from 15th century Germany of witches bringing down the rain Want: A weather prediction service that publicly states it is using Astrology, the I Ching and other divination techniques, to predict the weather. As an added bonus, it tells you about the implications of today's weather for your personal fortune. It would be about as effective as current weather prediction services, just far more fun and honest.
-
"Mordor gets gentrified with Starbucks cafes" - Prompt for VQGAN+CLIP Image generator
-
"Automation is an addiction. Stay in music school kids, play your instruments. Don’t do AI." - Quote from the new generative music album from Dadabots.
-
Alien Dreams: An Emerging Art Scene - by Charlie Snell
In recent months there has been a bit of an explosion in the AI generated art scene.
-
Multimodal Few-Shot Learning with Frozen Language Models (tweet)
Our new paper shows how to prompt a pre-trained text language model with a combination of text AND images. Keep the language model frozen and train a vision encoder to embed images into the same space as word sequences. The LM can generalize to accepting interleaved images and text in a prompt. This enables *multimodal* GPT3 style “in-context” few-shot learning. My favourite part about our system is you can access the factual “world knowledge” in the LM with visual input Conceptual Captions has named entities “hypernymed” away, so our model has not been trained to associate named entities with images of them.
-
Voice AI is scary good now. Video game actors hate it
A new ‘Witcher 3’ mod uses tech that’s ethically questionable and what one actor calls “utterly soulless.” But can anything be done about it?
#Comment: My talk "Automating Creativity Why?" (2018) and accompanying text seems more relevant than ever.
-
New works from Nvidia AI team (code & colab)
"It turns out CLIP can go beyond in-domain image editing - and actually convert a generator between domains! Here are some examples"
-
Computer scientists are questioning whether Alphabet’s DeepMind will ever make A.I. more human-like (CNBC) - I contributed a few comments to this piece, highlighted here. Received a few messages from peers asking "isn't this a bit too harsh?". Short Answer: No Sir, it isn't too harsh. Please don't take everything so serious - especially me, giant corporations, yourself, AI and God.
Not everyone is convinced, however. Samim Winiger, an AI researcher in Berlin, told CNBC that DeepMind’s “reward is enough” view is a “somewhat fringe philosophical position, misleadingly presented as hard science.”
He said the path to general AI is complex and that the scientific community is aware that there are countless challenges and known unknowns that “rightfully instill a sense of humility” in most researchers in the field and prevent them from making “grandiose, totalitarian statements” such as “RL is the final answer, all you need is reward.”
“In somewhat typical DeepMind fashion, they chose to make bold statements that grabs attention at all costs, over a more nuanced approach,” said Winiger. “This is more akin to politics than science.”
Winiger argues that we’re no closer to AGI today than we were several decades ago. “The only thing that has fundamentally changed since the 1950/60s, is that science-fiction is now a valid tool for giant corporations to confuse and mislead the public, journalists and shareholders,” he said.
-
Today is the 90th anniversary of Kurt Gödel's 1931 paper, which laid the foundations of theoretical computer science, identifying fundamental limitations of algorithmic theorem proving, computing, AI, logics, and math itself. - Text by Jürgen Schmidhuber
-
Had to leave a comment on (yet another) stealth ad video about self driving cars..
"While you US guys keep making sleek payed advertising clips for (self-driving) cars, the rest of the world had enough of cars, full stop. Self driving cars are still cars - dangerous, polluting, city destroying, loud metal boxes mostly filled with air - Highly inefficient legacy methods of transportation. A future worth aspiring to are cities free of all cars: Cities for people, not machines. The Europeans and even more so the Chinese have gotten the message, tripling down on high speed trains instead, which have a far superior space utilization profile than any car will ever have. Think bicycles, remote-working and walkable cities - not even more cars with just with some AI magic on-top which takes away the last few driver jobs (Uber & co) from people who are already struggling to make a living."
-
VQGAN+CLIP Prompt: "Moomin family portrait. realistic oil painting of Critters. unreal engine"
'Realistic oil painting of Mickey Mouse Pornography. unreal engine'
-
Detecting Conspiracy Theories on Social Media (Rand)
Google’s Jigsaw unit sponsors a RAND report that recommends infiltrating and subverting online conspiracy groups from within while planting authoritative messaging wherever possible. With a focus on online chatter relating to alien visitations, COVID-19 origins, white genocide, and anti-vaccination, the Google-sponsored RAND report published last week shows how machine learning can help detect and understand the language used by “conspiracy theorists.”
-
Synthetic living machines: A new window on life (cell)
Here, we review recent advances in the emerging field of synthetic morphogenesis, the bioengineering of novel multicellular living bodies. Emphasizing emergent self-organization, tissue-level guided self-assembly, and active functionality, this work is the essential next generation of synthetic biology. Aside from useful living machines for specific functions, the rational design and analysis of new, coherent anatomies will greatly increase our understanding of foundational questions in evolutionary developmental and cell biology.
