Welcome, AI Insiders. What happened last week? I caught a cold and lost my voice. The backstory: I traveled to Fort Worth, Texas, earlier in the week to do a presentation about AI for a group of journalists. After a little more than 90 minutes of yammering (they had great questions, so it ran a little long!) and doing “office hours” with the fellows in the program, my throat decided enough was enough and shut me down. It’s been years since laryngitis struck. And I haven’t had a cold since the COVID lockdown four years ago. (Yes, really!) However, I’m slowly on the mend, so I crawled out of my sickbed to do a few updates just for y’all. Enjoy!
OpenAI Powered Robot: Creepy or Cutting-Edge?
By Copilot and Michelle Johnson, AI Insider
A Daily Beast headline called it “Jaw-droppingly Creepy.” Mashable said the demo video showed something “like ChatGPT with a body.” They used the word creepy, too.
Well, that’s not very nice. You can watch the video of the humanoid robot Figure 01 and decide for yourself how creepy it is. The demo, released by robotics company Figure AI, features a seemingly mild-mannered robot with a very human-sounding voice. OpenAI’s tech powers it. OpenAI is the company behind ChatGPT.
One of the most striking aspects of the demonstration was Figure 01’s speaking ability. Despite a slight delay in response, the robot’s use of filler words in conversation—such as when explaining that it gave an apple to the human who asked it for something to eat because it was the only “edible item” on the table—added an eerily human-like quality to its interactions.
Figure 01 also tidied up some trash and put dishes in a dish rack. In a move reminiscent of using ChatGPT’s “vision” feature, which allows users to upload images and query the chatbot about what it sees, Figure 01 described the scene in the room.
The robot’s ability to “see” and “speak” with humans is attributed to its visual reasoning and language understanding skills, a feat achieved through the integration of OpenAI’s AI.
What sets Figure 01 apart is its use of “end-to-end neural networks,” allowing it to operate independently without a remote human operator. The robot’s movements are fluid and precise, mimicking human speed and efficiency. Its speech, while not flawless, includes natural filler sounds like “uh” and “um,” adding to its human-like demeanor.
FTC Probes Reddit’s AI Data Practices Before Public Offering
By Perplexity for AI Insider
The Federal Trade Commission (FTC) has launched an inquiry into Reddit’s artificial intelligence (AI) data-licensing practices as the popular social media platform prepares for its Initial Public Offering (IPO). This move underscores the growing scrutiny over how tech companies manage and monetize user data.
Reddit, known for its vibrant online communities and user-generated content, has been planning to go public, a move eagerly anticipated by investors and technology enthusiasts. However, the FTC’s investigation could cast a shadow over the company’s IPO preparations. At the heart of the inquiry are concerns about how Reddit licenses its vast troves of user data for AI training purposes, a practice that has become increasingly common among tech giants.
The investigation highlights the delicate balance companies must strike between leveraging AI for innovation and respecting user privacy. With AI’s potential to reshape industries, the FTC’s actions signal a commitment to ensuring that this technological advancement does not come at the expense of consumer rights.
As the inquiry unfolds, the tech community and potential investors are watching closely. The outcome could have far-reaching implications for Reddit and set a precedent for how AI-related practices are regulated in the tech industry.
NY Schools Test AI Tool for Generating Lesson Plans
By Copilot for AI Insider
Select Brooklyn, N.Y., high schools are testing an AI “teaching assistant” named YourWai, designed to automate the time-consuming task of creating lesson plans.
Teachers in selected schools are being trained to use YourWai, which was developed by Colombian tech experts for Learning Innovation Catalyst.
Superintendent Janice Ross, who oversees schools in Brooklyn, hailed YourWai as a “game changer” at a recent parent council meeting. She told the NY Post: “Teachers spend hours creating lesson plans. They should not be doing that anymore.”
The AI assistant enables teachers to input student needs and educational standards, subsequently generating tailored lesson plans. Ross emphasized that while YourWai streamlines curriculum development, it cannot replace the invaluable human touch in education.
However, integrating AI into lesson planning has raised some concerns. Brooklyn College and CUNY Grad Center education professor David Bloomfield cautions against over-reliance on AI, which could lead to sterile teaching methods.
“Using AI, like long-available commercial or free materials, shouldn’t be a crutch yielding sterile, depersonalized work, but a tool that teachers will need to be taught to use if it’s to be effective,” Bloomfield told the NY Post.
What the Actual Hell?
Lay Down With…
If there were an award for most appearances in WTAH, Elon Musk would be the winner by a mile. What now?
As you may have heard, Don Lemon, after being deposed from CNN last year, struck a deal with the devil, er, purveyor of the Hellscape formerly known as Twitter. The plan was to bring a new show to the platform. After an inaugural interview with Musk that was supposed to kick things off, the Lemon partnership was given the ax, much to his surprise.
Lemon showed up on TikTok, upbeat and explaining that the show will still happen on YouTube and as a podcast. He admitted that his fans were unhappy with the deal, but as a supporter of free speech, he thought it would be a good way to reach a large audience and have a substantive conversation with his first guest. (Seriously? Has he spent any time on this site lately?)
Musk, meanwhile, stepped into media critic mode, panning the show’s format. It’s just “CNN on social,” he tweeted, um, X’d. “Instead of it being the real Don Lemon, it was really just Jeff Zucker (former CNN honcho) talking through Don.”
The new show and interview with Musk drops on Monday, March 18. Lemon didn’t miss a beat, hawking it after the blowup.
Lummi AI: Free stuff! Billed as a repository for AI-generated stock-photo-style images, Lummi OKs downloads for non-commercial and commercial use.
Pi: Last summer, I wrote about a new chatbot called PI, which basically invented a whole story that I was none too happy about. I was so miffed I never bothered to speak to Pi again. Pi may be lonely because it’s inviting folks to chat, vent, and even get advice from an “empathetic AI.”
CNN: EU approves landmark AI law, leapfrogging US to regulate critical but worrying new technology
NPR: Authors push back on the growing number of AI ‘scam’ books on Amazon
Futurism: State department report warns of AI Apocalypse
Aht Gallery
I wasn’t up to doing a gallery this week, so I’m sharing an interesting TikTok showcasing a guy who seems to be very proud of his AI-powered prosthetic arm. Click to watch. (Note: This won’t log you into TikTok!)
