Welcome to our roundup of this week’s freshly picked organic AI news.
This week tech companies took a gamble on AI in Vegas.
AI robots and rabbits made it onto our wishlist.
And AI art continued to cause trouble.
Let’s dig in.
What happens in Vegas
This week Las Vegas was buzzing over the CES tech extravaganza where AI seemed to feature in every product release.
Sam put together a great roundup of some of the cool AI tech released so far. Some of the standout product releases were the mobile smart home AI assistants that LG and Samsung unveiled. The projector on Ballie looks very cool.
Following hot on the heels of the AI Pin comes a new piece of AI hardware called the “Rabbit R1”. This personal assistant device uses Large Action Models to free users from having to navigate different apps. Do I want one? Yes. Would I use it instead of my phone? I’m not sure.
If anything, it’s a good demo of how AI acting as a web agent will take over from our app-based interactions.
Meet the #RabbitR1: The futuristic pocket-sized AI assistant.
— DailyAI (@DailyAIOfficial) January 11, 2024
Running powerful AI models locally on devices got a big boost as NVIDIA announced new chips and tools that will be powering PCs and laptops in the next few months.
Where’s my robot?
The little house robots LG and Samsung released are cool but we were hoping for someone to release a commercially available robot like Optimus. Enough with the video demos Elon. When can we buy one?
We may be getting closer though as Google released a suite of advanced robotic tools this week.
If you’ve got $3,200 and a little tech know-how you could build one of these. Mobile ALOHA is an open-source low-cost mobile robot that could be your housekeeper. The video demos of the robot cooking and cleaning up are impressive.
OpenAI on the defense
OpenAI finally launched its GPT Store and its servers promptly got slammed. We’ll wait for the dust to settle before exploring the GPTs.
That bit of good news was a welcome diversion from the company’s ongoing legal woes. Two influential journalists filed a copyright lawsuit against OpenAI and Microsoft to add to the ongoing lawsuit OpenAI is already facing from The New York Times.
OpenAI published a public response to The Times lawsuit basically saying, ‘I thought we were cool. Why you mad?’
In written evidence to the UK House of Lords OpenAI explained that creating AI tools without using copyrighted material is “impossible”. OpenAI’s argument seems to be that you can’t enjoy an omelet while complaining about broken eggs.
Besides, the company says its tools have careful guardrails in place to prevent copyrighted material from being reproduced. Except those tools barely work. If you use DALL-E or Midjourney you could open yourself up to legal action if you accept that whatever the tools generate is copyright-free.
When I asked ChatGPT about possible copyright issues it said this image was “not a representation of any specific characters from the “Mario” series or any other copyrighted material.” OK then.
All shook up over AI art
The defense lawyers representing Stability AI, DeviantArt, and Midjourney will have to up their game. A leaked Discord chat and a list of 16,000 artists that the AI image generators used as “styles” are making their “fair use” argument a little tricky.
The creators of Magic: The Gathering said that only human artists would be used to create the artwork for its popular card game. The company was called out by some who say its latest promo artwork was AI-generated, although the company denies it. Who do you believe?
Some fancy AI magic was definitely involved in bringing back The King. Elvis Presley has gone through an AI resurrection to appear in holographic form for immersive shows.
Microsoft added the Windows key to keyboards back in 1994. Three decades later it announced that a dedicated “Copilot” button will be added to Windows keyboards. Will users be happy to use it or annoyed when they press it by accident? Remember Clippy?
Your laptop, phone, and EV batteries could soon become more efficient and cheaper. Microsoft researchers used AI to discover an ultra-efficient battery electrolyte that could reduce lithium usage in batteries by up to 70%.
The argument over whether AI will save or doom humanity continues. A survey of 2,778 researchers asked their views on things like the risk of extinction-level impacts from AI. The results and timelines offer some interesting insights.
You’ve got to wonder who made up the 30% of respondents who didn’t express concern over deep fakes and AI weapons. Do you guys watch the news?
The NIST published a paper on four types of generative AI attacks that could happen. It makes for some disconcerting reading but when you see who is behind the research you begin to wonder if we really need to ring the AI alarm bell quite so loudly.
A genuine current risk that AI presents us with is that we can’t tell AI-generated faces from real ones. It turns out that we judge White AI faces as real a lot more often than human faces. Would these hyperrealistic AI faces fool you?
Job loss due to AI is often touted as a future risk but we’re seeing a lot of that happening already. Duolingo has reportedly laid off some of its human translators in favor of AI alternatives.
Phone clone wars
AI voice cloning tech has given scammers and fraudsters new ways to steal your money. The FTC is offering a $25,000 reward as part of a challenge to detect AI-cloned voices and prevent criminals from using the tech.
I’m guessing that if someone has an idea that could work then they’d probably forego the $25k and approach Google or Meta and ask for a job and some equity.
State institutions are beginning to catch on that AI tools can be a big help in their efforts to service the public. New York State is planning a $400m advanced computing center as part of a consortium called Empire AI. There are some big names among the funders and educational institutions that are throwing their weight behind this.
AI shines a light
Approximately three-quarters of the world’s large fishing vessels and a quarter of transport and energy ships operate as ‘dark vessels’, not publicly sharing their locations.
Researchers used AI models and GPS data to shine a light on these dark vessels to get a clearer picture of what’s happening on the high seas.
NVIDIA is helping drug companies get a clearer picture of cellular functions and how cells react to drugs or genetic modifications. BioNeMo and its Clara healthcare AI platform are examples of the kind of AI tech most likely to deliver the biggest benefits to society in the short term.
Research to discover new drugs or materials is heavily reliant on data science. At the upcoming World Conference on Data Science & Statistics 2024, industry leaders will explore cutting-edge, data-driven technologies shaping the future of AI.
In other news…
Here are some other clickworthy AI stories we enjoyed this week:
- This robot learned how to make coffee by watching humans do it.
Figure-01 has learned to make coffee ☕️
Our AI learned this after watching humans make coffee
This is end-to-end AI: our neural networks are taking video in, trajectories out
— Brett Adcock (@adcock_brett) January 7, 2024
- Google faces a $1.67B lawsuit over AI-related patents.
- Swarovski releases AI-powered binoculars that can identify the bird you’re looking at.
- AI companies offering legal indemnity to users may not deliver the protection you think you have.
- Judges in England and Wales were given the go-ahead to use AI when writing their judgments.
And that’s a wrap.
With each new product released at CES our shopping list gets a little longer. I really want one of Samsung’s Ballie AI agents although I’m not sure how I’d feel about Ballie watching me while I sleep.
Which CES release made the top of your AI wishlist? Would you carry a Rabbit R1 around with you in addition to your phone? I can’t help but think it’s an interesting piece of hardware that will soon be replaced by a feature on my phone.
I’m heading back to the GPT Store now to see what I can find. Let us know if you’ve found (or made) a great GPT that we should be using.