AI & Robotics News and Discussions

[firestar464]

https://techxplore.com/news/2024-03-acc ... uards.html

Many publicly accessible AI assistants lack adequate safeguards to prevent mass health disinformation, warn experts

[weatheriscool]

Unitree H1 is first humanoid to nail a backflip without hydraulics
By Paul Ridden
March 21, 2024

After setting a new world speed record for humanoid robots earlier this month, China's Unitree is now claiming another. Its latest H1 bipedal takes the title for first to perform a standing backflip without the use of hydraulics.

Yes, humanoids like Boston Dynamics' Atlas have been nailing backflips for a few years now but they make use of heavy, potentially leaky hydraulics to launch into the air, somersault backwards and then land on both feet.

Impressively, Unitree's H1 relies on in-house M107 electric joint motors only, each of which boasts 360 Nm (265.5 lb.ft) of peak torque and can also be found on the company's B2 quadruped. Each leg has three degrees of freedom at the hip plus one at the knee and another at the ankle, and all cabling is routed internally for snag-free clean lines.

Unitree used reinforcement learning simulation to train the H1 in the art of in-place flipping, which it almost pulled off as planned – save for a corrective mini jump at the end.

https://newatlas.com/robotics/unitree-h1-backflip/

[firestar464]

https://techxplore.com/news/2024-03-sci ... milar.html

Top computer scientists say the future of artificial intelligence is similar to that of Star Trek

[firestar464]

https://techxplore.com/news/2024-03-mac ... right.html

Machine 'unlearning' helps generative AI forget copyright-protected and violent content

[weatheriscool]

NVIDIA to create AI ‘agents’ that outperform human nurses
By Paul McClure
March 25, 2024

NVIDIA has partnered with Hippocratic AI to develop AI-powered ‘healthcare agents’ that have already been shown to outperform other large language models and human nurses in specific tasks. Will these agents help ease the global healthcare worker shortage, or will they lead to more problems?

Workforce shortages are simply an imbalance between need and supply. In healthcare, this equates to not being able to provide the right number of people with the right skills in the right places to provide the right services to the right people.

The World Health Organization (WHO) has estimated a projected shortfall of 10 million health workers by 2030, mostly in low- and middle-income countries. However, the pinch produced by shortages is already being felt in rural and remote settings in high-income countries like the US and Australia. Ostensibly addressing the global healthcare staffing crisis, NVIDIA recently announced their partnership with Hippocratic AI to develop generative AI-powered ‘healthcare agents’.

“With generative AI, we have the opportunity to address some of the most pressing needs of the healthcare industry,” said Munjal Shah, cofounder and CEO of Hippocratic AI. “We can help mitigate widespread staffing shortages and increase access to high-quality care – all while improving outcomes for patients. NIVIDIA’s technology stack is critical to achieving the conversational speed and fluidity necessary for patients to naturally build an emotional connection with Hippocratic’s Generative AI Healthcare Agents.”

https://newatlas.com/technology/nvidia- ... ai-nurses/

[Time_Traveller]

Future of AI will be like the Borg in Star Trek, experts predict

3 days ago



Different artificial intelligence units will one day be able to team up and share information with each other – just like the Borg in Star Trek, according to leading computer experts.

The Borg are cybernetic organisms in the sci-fi TV show which operate through a linked hive-mind known as “The Collective”.

Scientists from the universities of Loughborough, Yale and the Massachusetts Institute of Technology have said humanity is set to see the emergence of “Collective AI”, where many different units – each capable of continuously learning and gaining new skills – form a network to share information.

The team unveiled their vision in the journal Nature Machine Intelligence.

But the researchers added that unlike the antagonists from the Star Trek franchise or the villainous Replicators – who are a highly advanced machine race in the sci-fi series Stargate SG-1, they expect the impact of Collective AI to be more positive.

https://www.independent.co.uk/news/scie ... 17180.html

[firestar464]

https://phys.org/news/2024-03-scientist ... istry.html

Scientists deliver quantum algorithm to develop new materials and chemistry

[weatheriscool]

Robotic face makes eye contact, uses AI to anticipate and replicate a person's smile before it occurs
https://techxplore.com/news/2024-03-rob ... icate.html
by Holly Evarts, Columbia University School of Engineering and Applied Science

What would you do if you walked up to a robot with a human-like head and it smiled at you first? You'd likely smile back and perhaps feel the two of you were genuinely interacting. But how does a robot know how to do this? Or a better question, how does it know to get you to smile back?

While we're getting accustomed to robots that are adept at verbal communication, thanks in part to advancements in large language models like ChatGPT, their nonverbal communication skills, especially facial expressions, have lagged far behind. Designing a robot that can not only make a wide range of facial expressions but also know when to use them has been a daunting task.
Tackling the challenge

[funkervogt]

This excellent essay on Astral Codex Ten explains why the success of GPT-5 is so important to the pace of AI development. Let me explain:

There have been four GPT AI's so far: GPT-1, 2, 3, and 4. On average, each took 18 months to invent after the last.

Each one cost roughly 30x as much money to create as the last. Because GPT-4 cost $100 million, that means GPT-5 will cost about $2.5 billion.

The company that created the GPT machines, OpenAI, has enough money to make GPT-5, and it will probably be released in the next year. However, the cost trend means GPT-6 will cost $75 billion, which is well beyond its means.

Microsoft owns OpenAI, and if GPT-5 is a blockbuster product (e.g. - eerily smart and competent), Microsoft could, with difficulty, make the $75 billion investment to build GPT-6.

The cost trend also reveals that GPT-7 will cost $2 trillion to build, which no one has to spare. Only a major country or group of countries could afford it, and they'd only spend that much if GPT-6 were a super-blockbuster product, making the reward of an even better computer worth it.

The amount of computer chips and electricity needed to support each successive GPT machine has also been increasing exponentially, and while building GPT-5 is still within OpenAI's means, GPT-6 isn't--it would need something like a town-sized server farm and a dedicated nuclear reactor.

So if GPT-5 is underwhelming, isn't massively adopted across the world for all kinds of tasks, and doesn't make OpenAI billions of dollars in profit per year, it won't be worth it for anyone to build GPT-6 at the same pace the previous five GPTs were built. Instead, they'll slow down and build GPT-5.1, then -5.2, etc., with a year or two between each iteration. Moore's Law and variety of small hardware and software improvements will make a slower development strategy affordable: by the time GPT-6 capability is achieved, the total cost will have been much less than the $75 billion that would have been needed to build it a decade sooner and in one, big leap.

If the scenario in the preceding paragraph materializes, the general public will misperceive it as "proof" that AI was hype and can't actually be built, in the same way that the slowdown in progress with autonomous vehicles has convinced people the technology will never succeed.

The scaling problem with LLMs hits home how bad their learning algorithms are, and how much the tech industry is trying to brute force its way to creating an AI by throwing more computer chips and electricity at the problem. Within the next three or four years, we'll know whether this paradigm can continue, and there will be major implications for the tech industry and popular perceptions of "AI."

If GPT-5 is really good, then buckle up because mass job displacements, social changes, science and tech acceleration, and Manhattan Project-levels of investment in AI that will be felt across politics and the world economy will start before the end of this decade.

However, I predict it's much likelier GPT-5 will fall below the necessary performance threshold. While an impressive product, it won't be good enough to inspire the immediate creation of GPT-6. LLM's will probably suffer from an insurmountable "last mile problem" that keeps them from evolving into true AGIs, though in many narrow domains they will equal or surpass human intelligence and competency and they'll be able to convincingly "fake" general intelligence 99% of the time (which some dumber humans will mistake for proof GPT-5 is sentient). I believe we'll have to wait for several important algorithmic breakthroughs to happen to achieve true AGI, which I think could happen as early as 2050.

https://www.astralcodexten.com/p/sam-al ... 7-trillion

I have an important update. Re-read the whole quoted passage, but pay special attention to the sentences I underlined.

News has leaked that Microsoft has agreed to pay up to $100 billion to build massive data centers in the U.S. to support OpenAI's future work. Codenamed "Project Stargate," the data centers will be operational in 2028.
https://www.reuters.com/technology/micr ... 024-03-29/

If you open the link at the bottom of my quoted passage, you'll find a line chart that estimates GPT-6 will be unveiled in 2027 or 2028 if the current pace of work keeps up.

I think GPT-5 was secretly created months ago, OpenAI showed it to Microsoft's leadership behind closed doors, and Microsoft was so impressed with GPT-5 potential as a commercial product that they've determined its successor, GPT-6, will be worth more than $100 billion in profit.

Reading the tea leaves, I think this prediction of mine might be wrong.

If GPT-5 is really good, then buckle up because mass job displacements, social changes, science and tech acceleration, and Manhattan Project-levels of investment in AI that will be felt across politics and the world economy will start before the end of this decade.

However, I predict it's much likelier GPT-5 will fall below the necessary performance threshold. While an impressive product, it won't be good enough to inspire the immediate creation of GPT-6.

[firestar464]

Didn't sources say that they were gonna release GPT-5 in the middle of this year?