Kid Klava wanted to sing a song he wrote, but realized his chops really weren’t up for the task. So why not get John Lennon sing it – with Paul on backing vocals? He claims it was surprisingly easy and it only took a few minutes to generate. If you’d like to have a go yourself, there are dozens of great YouTube tutorials.
Have you ever heard of a movie about an advanced AI based defense system that becomes sentient? After being handed full control, it expands on its original nuclear defense directives to assume total control of the world despite its creators’ orders to stop.
It’s not War Games or Terminator. Nope, this movie is from 1970 and it’s called Colossus: The Forbin Project and it’s available on Vimeo for free. Maybe give it a watch before it becomes reality.
Who would think something like improving a chip design tool would cause the International Symposium on Physical Design (ISPD) to get so out of hand it was called a “trainwreck” and an “ambush” of the presenters.
The crux of the clash was whether Google’s AI chip design layout solver was really better than those of humans or state-of-the-art algorithms. In the end, there was a firing and a wrongful termination lawsuit with Google, a team re-doing a years worth of work, ugly accusations and drama, and 2 AI researchers leaving a promising field of AI improved chip design due to the conflict.
The argument involved a lot of factors like comparing different chip placement algorithms, results of reinforced learning, initial placement bias, metrics of success such as wire length, annealing, and general benchmarks – but in the end resulted in a circus of accusations, lawsuits, and drama.
Read more about it here on the IEEE article
Combining live voice recognition, voice synthesis, and an AI large language model, Replica Studios’ Smart NPCs for Unreal allows you to carry out real-time, non-scripted conversations with anyone on your virtual streets. You can even download the demo here and try it yourself (warning: it’s 25 gb!).
These fully simulated AI characters will respond to your live-spoken language in the demo. You can ask where they’re from, whether they believe in aliens, complement their clothes, what car they drive, get food recommendations, tell people you don’t like their haircut, ask for directions to the park (and they’ll give it to you and tell you to feed the pigeons), talk about their insurance plans, what crimes they have committed, start some rumors and conspiracy theories, and anything you can think of. It’s got a few quirky bits, but it’s pretty surreal.
If you think that’s crazy, also check out their AI powered voice actors. The amount of dynamic quality in their voices is pretty amazing. It is almost good enough to replace main character voice actors, but it sure could be used to fill out dozens and dozens of minor roles.
The original channel comes from Jeff Favignano:
The 15-second audio clip sounds like a muffled version of Pink Floyd’s Another Brick in the Wall played underwater. Except Pink Floyd didn’t perform any of the music in the clip. Instead, the track was captured by a team of researchers at the University of California at Berkeley, who looked at the brain activity of more than two dozen people who listened to the song.
That data was then decoded by a machine learning model and reconstructed into audio — marking the first time researchers have been able to re-create a song from neural signals.
Articles:
Another blow to human superiority?
ChatGPT 4 can already pass the Bar Exam, now it appears to be more creative than humans too. Is this a blow to arguments that computers can’t replace artists?
A University of Montana study used ChatGPT 4 take the Torrance Test of Creative Thinking, a well-known tool used for decades to assess human creativity. They then submitted the AI’s test answers along with the test taken by 24 UM students taking Dr. Erik Guzik’s entrepreneurship and personal finance classes.
These scores were then compared with 2,700 college students nationally who took the TTCT in 2016. All submissions were scored by Scholastic Testing Service, which didn’t know one of the submissions was done by ChatGPT.
The result? The AI’s test answers placed in the top percentile for fluency – the ability to generate a large volume of ideas – and for originality – the ability to come up with new ideas. The AI slipped a bit – to the 97th percentile – for flexibility, the ability to generate different types and categories of ideas.
“For ChatGPT and GPT-4, we showed for the first time that it performs in the top 1% for originality,” Guzik said. “That was new.”
Links:
FoxMaster uses a wide variety of AI tools: vision, object recognition, chatgpt, and others to give Laura Croft the AI treatment. She not only can traverse the game, but also has personality and narrates what is going on.
Foxmaster admits some of this is not complete and may be stretched a bit – but his analysis and breakdown of the problems of navigation, identification, and character personality into discrete problems is very interesting.
Beginning early July, mirrored spheres began popping up in cities across the world. It is not public art but a 6.2 pound biometric imaging device designed to scan your eyeballs and capture your irises.
No, it’s not a joke. The company doing this is Worldcoin. Worldcoin was founded three years ago by Alex Blania and Sam Altman, CEO of OpenAI. Their intent is to create “a new identity and financial network connecting billions of people in the age of A.I.” via a privacy-ensuring digital identity it calls World ID and a digital currency, WLD. World ID is supposed to be a ‘perfectly safe’ global identity protocol to enable individuals to prove their personhood online in an era of rampant A.I. deepfakes. WLD is a tool to build an “A.I.-funded UBI [universal basic income]”
This should all sound familiar because they’re the same arguments being made for digital currencies like Bitcoin. The company stresses they have security all along the trust chain, but it reads more like a dystopian nightmare in which everyone has been cataloged and identified by an unknown party with unknown motives. This is a company that could be working for anyone. A 3rd party agent that has unknown motives, unknown technical expertise, and unknown longevity to keep your biometric data safe. You do not really know which government, people, nor company is really behind it all nor what their values nor legal protections you would have. It’s a terrify black box of giving up your biometric data to an almost completely unknown entity. I vote a hard no.
In my opinion, this is a violation of privacy and an extremely bad idea. Even if they are secure today – are they really ready to protect your biometric data, collected surreptitiously, without consent, for all time, and never to sell it? It seems pretty unlikely as every country and every company who has promised this before hasn’t lasted 10 years before being hacked, leaking, being forced to turn over that data, or just flat selling you and your data to the highest bidder when they’re bought out or go bankrupt. Ready for North Korea, China, or Putin to have access to your World ID and any money you put in WLD?
So, maybe be sure to wear some good sunglasses when you find a shiny orb laying on the street.
At least some folks are starting to do an investigation – one I suspect will end badly for Worldcoin considering the EU’s less than open stance on collection/storage of biometric data collection and privacy.
AI is here. Netflix put out a short show called Unknown: Killer Robots that is largely on-point. If anything, it’s not as up to date as reality – which is a little scary.
Despite our efforts, there’s really no putting the AI genie back in the bottle despite the attempts of artists, politicians, and academic pundits. AI has demonstrated it can teach robots how to walk and fly better than any static system, it can create art faster and with nearly the same quality as real artists. The question is, can we control and limit it in a reasonable way – or will it destroy us?
To that point, the show demonstrates how countries are increasingly seeing AI in military use. We’re already seeing off-the-shelf drones being used extensively in the Ukrainian war – from tiny drones scouting and delivering hand grenades to ad-hoc drones using small plane engines to delivering bombs. They also cover the counter-intuitive reality that trying to save lives by developing military technology has almost always lead to even greater casualties – such as the Gatling gun which was design to put fewer people on the military field but resulted in massive causalities of mis-matched soldiers in WW1
Many of the topics and data they cover in this show are actually old news – the reality is that AI enabled systems are around 5-7 years more advanced. Which is almost half the length of time the modern field has existed.
Still, they cover most major bases:
The stakes couldn’t be higher. While politicians can argue the ethics, the reality is that when forces are pushed to their breaking point and a force is about to lose – just about nothing is off the table. Especially ad-hoc and terrorist forces which have perpetrated chemical, biological, and conventional weapons attacks (from bombings to shootings).
Some of the better quotes:
How could this go wrong? Smart as ChatGPT and fully autonomous.
Robotics and AI enthusiasts at the YouTube channel Underfitted recently upgraded their Boston Dynamics Spot quadruped robots with ChatGPT.