e497 — Rise of the Robots

Apple Image Playground generated image: robot holding a sock
Apple Image Playground generated image: robot holding a sock, January 2025

Published 13 January 2025

e497 with Michael, Michael and Andy on , , , , , , , and a whole lot more!

Michael, Michael and Andy are back in full co-host force once again, covering the robotics-filled CES innovations hitting the feeds this week.  

The team starts things off this episode discussing a TechCrunch article covering Nvidia’s R2X avatar.  The user experience for this avatar is through the computer desktop with a humanlike embodiment to provide assistance to users, seeing what they see on the desktop, or ingesting documents.  This reminds Michael R of the Reallusion Crazytalk capability (discussed way back in e35 in 2013), Andy of Microsoft’s Clippy, and Michael M of the Talking Moose.  The R2X puts “a face on agentic AI” according to Nvidia’s press release, and it is easy to imagine how Clippy, the Talking Moose or other embodiments might be a similar face.  

Moving from virtual to physical, it is clear that CES2025 is chock full of robots, for all manner of purposes.  

The uncanny valley resurfaces with the Realbotix humanoid robot Melody. The facial and body movements of the robot resemble human natural movement, but are still far enough away to create the emotional discrepancy described by the uncanny valley.  In addition, Andy makes the point that the reported open source nature of the robot and its code does not appear to be substantiated.

Moving along from the humanoid form factor to other robots, there are many robot examples whose purpose is to vacuum, mop, deliver drinks or even pick up socks.  Check out the show notes below for some of the CES robots, from those that attach to purses, to those that cool your tea.  

Michael R shares a project he’s been working on dealing with language localization – check out the link below for more details.

Wrapping things up for the episode, the co-hosts turn to a couple LEGO innovations: the latest in the Nintendo-LEGO partnership, and an intriguing Hackaday omni-directional treadmill.

What tasks would you want a robot to do for you?  Have your bots 🤖 drop our bots 🤖 a line at @gamesatwork_biz (our home for now) and let us know! 

These show notes were lovingly hand crafted by a real human, and not by a bot.  All rights reserved.  That’s our story and we’re sticking to it.

Selected Links

AI

TechCrunch article: Nvidia’s AI avatar sat on my computer screen and weirded me out

Nvidia press release 6 Jan 2025: NVIDIA Launches AI Foundation Models for RTX AI PCs

Reallusion’s Crazytalk

Games at Work e35: Pirates of Pizzazz

Wikipedia article: Office Assistant (Microsoft Clippy)

Wikipedia article: Talking Moose

Uli’s Moose

Robots, Robots, Everywhere

Interesting Engineering article: Melody: New humanoid robot companion that packs into a suitcase unveiled

Realbotix

Wikipedia entry: Uncanny Valley

The Hitchhiker’s Wiki article: Zaphod Beeblebrox

9 to 5 Mac article: These are the silliest robot cleaners at CES – including one that carries drinks

TechRadar article: This furry clip-on robot is the strangest thing I’ve seen at this or any CES

TechRadar article: Roborock’s new robovac has a mechanical arm that can pick up your socks and maybe also play with your cat

Roborock

The Verge article: SwitchBot is bringing a Rosie the Robot wannabe to CES

CES article: Stop Scalding Your Tongue: This $25 Cat Robot ‘Blows’ on Your Drink to Cool It

Yukai Engineering: Yukai Engineering to Debut ‘Nékojita FuFu,’Attachable Mini-Robot That Cools Hot Drinks/Food, at CES Unveiled

Leveraging Language Localization

Michael Rowe’s Random Thoughts blog post: Coming Soon – Quick Localizer

LEGO

Brickfanatics article: LEGO Nintendo Game Boy officially announced for 2025

Hackaday article: Gaze Upon This Omni-directional Treadmill’s Clever LEGO Construction

Games at Work e412: 3D or not 3D – Virtuix Omni treadmill

Games at Work e451: Fahrenheit – Disney Holotile treadmill

E496 – The Starting Square

Photo by Benjamin Bousquet on Unsplash

Andy and Michael R are checking out the upcoming FOSDEM 2025 conference (Free and Open Source Developers European Meeting). This got them thinking about Episode 494 – License to Brick, where they talked about how the makers of the Moxie emotional support toy for kids were shutting down. But guess what? The CEO decided to let a team of developers create an Open Source version of the server and make it available for everyone!

After both of our hosts find the same Toot on Mastodon, about how a town in Poland is using Clams to monitor their water supply. While the story is not new, there is a very interesting documentary available for free on Youtube. This discussion takes us into the world of home automation and brilliant example – Operation Toilet!

A quick pivot takes us to a tool that has been created to allow Open Street Map data to generate MineCraft worlds. Perhaps this new world could be used to test out the new VR glasses that scientists have created for mice. Andy and Michael spend time enjoying the sentences in the article, as it appears that there may be a whole industry of “standard mouse VR”. Is this really common? Really?

We wrap up the show with a few preannouncements from LG for the upcoming CES. It appears that they will continue to add more and more screens in the house. We are sure the main use case is for new ads. We can’t wait to see what else is announced as CES.

Show Links

Moxie works on open sourcing their servers – https://www.techdirt.com/2024/12/30/embodied-is-actually-trying-to-release-moxie-robots-to-the-open-source-community/ 

Poland city uses 8 clams to secure their water supply – https://hear-me.social/@tylerknowsnothing/113743591435484939

Generate MineCraft Replicas based on Open Street Map – https://www.tomshardware.com/video-games/pc-gaming/minecraft-tool-lets-you-create-scale-replicas-of-real-world-locations-arnis-uses-geospatial-data-from-openstreetmap-to-generate-minecraft-maps 

VR Goggles for Mice – https://gizmodo.com/scientists-built-tiny-vr-goggles-for-mice-2000543775 

LG Announces microwave with 27 inch screen – https://www.theverge.com/2024/12/30/24331994/lg-microwave-27-inch-display-speakers 

LG Announces new monitors for gaming – https://www.theverge.com/2024/12/29/24331748/lg-ultragear-gx9-bendable-oled-5k2k-monitors-specs 

Our friend Roo’s Connected Fridge Tumblr – https://www.tumblr.com/fuckyeahinternetfridge 

e495 — Personal Planetarium

Inscription on the sundial at the Morehead Planetarium reading “When it is noon in Chapel Hill, it is 12:00 noon New York, 11:00am Chicago, 10:00am Denver, 9:00am San Francisco, 7:00am Honolulu, 6:00pm Oslo, Stockholm, Berlin, Amsterdam, Paris, Vienna, Rome, 5:00pm Dublin, London, Lisbon, 2:00pm Rio de Janeiro, 8:00pm Moscow, 10:30pm Calcutta, next day 2:00am Tokyo, 1:00am Hong Kong”
Photo by Michael Martine, Morehead Planetarium Sundial, Chapel Hill Sept 2019

Published 30 December 2024

e494 with Michael and Michael on personal planetariums, digital doppelgängers, agentic AI acceleration, and a whole lot more!

Michael and Michael start this episode off with a full dome sized bang, Sandwich Vision’s Theater app.  In addition to replicating famous theater spaces via Apple Vision Pro, Theater 2.0 also introduces a planetarium experience.  Michael R makes a North Carolina tie to the Morehead Planetarium and Science Center in Chapel Hill.  Michael M is reminded how the planetarium VR experience has surfaced on the podcast before.  See the link in the show notes to e177 from 2017 for one such example.

The team calls back to last week’s episode, dealing with AI discovering the rules to games thousands of years old with a new story from Nature.  This article explores the use of AI in translating animal communications from birds, whales, elephants and more, as well as identifying unique animals.  Additional use case for AI and GenAI include Apple’s Image Wand in iOS & iPadOS 18.2, a physics engine called Genesis, and creating a meeting doppelgänger via HeyGen.  The idea of sending an agentic version of yourself to a meeting sparks a lively discussion about responsible AI and what this bodes for the differentiation that the human may bring.  The trustworthiness (truthiness?) of chatbot responses surfaces via a pair of stories where people say “I asked ChatGPT …” or how the chatbot itself tends to follow mechanics similar to how mentalists interact with their audience.   

Wrapping things up for the episode, Michael and Michael anticipate the upcoming CES with a couple of hardware stories: a supremely expensive transparent television and a MagSafe gaming controller.

If you could talk with the animals, what would you say?  What new consumer electronics do you hope to see from CES 2025?  Have your bots 🤖 drop our bots 🤖 a line at @gamesatwork_biz (our home for now) and let us know! 

These show notes were lovingly hand crafted by a real human, and not by a bot.  All rights reserved.  That’s our story and we’re sticking to it.

Selected Links

A VRy Personal Planetarium

9 to 5 Mac article: Apple Vision Pro just got a planetarium, and it’s friggin’ awesome

Sandwich Vision‘s Theater: Cinema & Events app for Apple Vision Pro

IMAX app for iPhone, iPad and Apple Vision Pro

Reef Distribution 

Morehead Planetarium and Science Center

Games at Work e177: What could possibly go wrong?

Adler Planetarium

AI

Nature article: AI decodes the calls of the wild

IMDb: Dr. Dolittle, 1967

CNet article: Conjure Drawings From Sketches Using Image Wand in Apple Intelligence on Your iPhone or iPad

GitHub: Genesis: A Generative and Universal Physics Engine for Robotics and Beyond

Globe and Mail article: I created an AI avatar of myself to go to Zoom meetings

HeyGen

The Verge article: Here’s a new way to lose an argument online: the appeal to AI

SoftwareCrisis.dev article: The LLMentalist Effect: how chat-based Large Language Models replicate the mechanisms of a psychic’s con

National Geographic article: How technology is reshaping religion

Marie Kondo – KonMari

Browsing the past, and the future (CES 2025)

Tom’s Hardware article: Relive surfing the original internet with new emulator — 34 years later, WorldWideWeb app commemorates the first web browser

LG product page: 77 Inch Class LG SIGNATURE OLED T: World’s first Transparent 4K Smart TV 2024 with True Wireless Video & Audio Transfer

NotebookCheck article: M-Con: New MagSafe gaming controller for smartphones revealed before CES 2025 with Hall effect joysticks and pocketable design

Wikipedia article: Danger Hiptop

e494 — License to Brick

North Carolina First in Flight automobile license plate reading BRICKED
North Carolina custom automobile license plate “BRICKED”

Published 23 December 2024

e493 with Michael, Michael and Andy on digital storage for a century, AI datasets, videos & new Oreo flavors, hacking digital license plates and robots, and a whole lot more!

Andy, Michael and Michael start this episode off with a detailed article from the Harvard Law School Library Innovation Lab detailing the challenges and complexities of storing and retrieving digital data for a hundred years.  Sticking with the same institution of higher education (and the same Law Library), the next discussion deals with a dataset of one million public domain books to give the general public “highly-refined and curated content” to build AI models per the Wired article.  Next up are discussions on the use of AI in generating video content and new Oreo recipes.

Then, the co-hosts turn to an article on hacking the Reviver digital license plates in use in California and Arizona.  The team predicted that this may happen in 2018 in episode 200, in a very ‘back to the future’ moment.  

Hardware supported by cloud services and data are always susceptible to those services continuing to be available.  When the companies supporting the data and services close down, the hardware is bricked.  A story for this week about emotional support robots for children becoming bricks reminds the co-hosts of the Gatebox anime-style holographic AI from episodes 159 and 218.

Wrapping things up for the episode, before they are ‘OUTATIME”, the co-hosts discuss how AI can be used to recreate the rules from thousand year old games and congratulate the newest chess world champion.

What data would you want to store for more than a century?  Have your bots 🤖 drop our bots 🤖 a line at @gamesatwork_biz (our home for now) and let us know! 

These show notes were lovingly hand crafted by a real human, and not by a bot.  All rights reserved.  That’s our story and we’re sticking to it.

Selected Links

Data Storage

Harvard Law School Library Innovation Lab Century-Scale Storage

The Long Now Foundation

National Park Service History of Muir Woods

Wikipedia article: Disk pack

AI

Wired article: Harvard Is Releasing a Massive Free AI Training Dataset Funded by OpenAI and Microsoft

Institutional Data Initiative at Harvard Law School Library

MIT Technology Review article: This is where the data to build AI comes from

Ars Technica article: A new, uncensored AI video model may spark a new AI hobbyist movement

Gizmodo article: Oreo Maker Says It’s Using AI to Create New Snacks

Hackin’ like 007

Wired article: Hackers Can Jailbreak Digital License Plates to Make Others Pay Their Tolls and Tickets

Reviver

Games at Work e200: Eye in the Sky

Gizmodo article: This Playdate Mod Turns the Handheld Into the Cutest Little Robot Ever

Techdirt article: Startups Implosion Will Render $800 Emotional Support Robots For Children Into Useless Bricks

Games at Work e218: Virtually Married

Games at Work e444: Glitch in the Matrix

Thousand Year Checkmate

New Scientist article: The ancient board games we finally know how to play – thanks to AI

David Allen Green blog post: “Twelfth Night Till Candlemas” – the story of a forty-year book-quest and of its remarkable ending

Music Genome Project

GameTable article: Computational Techniques for Tabletop Games Heritage

chess.com article: 18-Year-Old Gukesh Becomes Youngest-Ever Undisputed Chess World Champion