Intelligent Machines (Audio) – IM 852: Gluten-Free Slop – AI at CES 2026
Host: TWIT
Panelists: Leo Laporte, Jeff Jarvis, Paris Martineau
Guest: Jason Hiner (Editor-in-Chief, The Deep View)
Date: January 8, 2026
“I think 2026 is going to be a very interesting year in AI.” — Leo Laporte (00:55)
Episode Overview
In the first Intelligent Machines episode of 2026, the panel takes a deep dive into the Consumer Electronics Show (CES) in Las Vegas, examining how artificial intelligence is shaping not only the year ahead but redefining the very essence of the event. With Jason Hiner joining live from the airport lounge after a week on the ground at CES, the conversation centers on what’s real and what’s hype in AI, the industry’s new obsession with “physical AI” and robots, notable product launches and trends, and larger implications for consumers, enterprise, and society.
Key Topics and Discussion Points
1. CES 2026: The AI Show
-
A New CES – Not Just for Consumers Anymore
- Attendance back up ("140,000 people, 4,000 vendors, 5,000 press," (03:42)), but the show feels increasingly enterprise/industry focused.
- Keynotes and main exhibits largely targeted at business, not average consumers.
- Leo: “Isn’t that strange? It’s not a consumer electronics show fully anymore. Maybe that’s why they took away the name.” (04:29)
- Jason: "It certainly feels like it’s on the upswing again." (03:42)
-
AI Hype Overload
- AI referenced everywhere (“How many times do you think you heard AI in a 24 hour period?” Paris Martineau (06:24)).
- Jason describes a mission to sift out "real AI products" from "AI washing." (06:35)
2. Nvidia’s Announcements: The Vera Rubin Platform & Beyond
-
Jensen Huang’s Keynote – Communicating Vision
- Praised as an “amazing educator” who communicates complex advancements ("he would...impact a lot of people," Jason, 07:39).
- New Vera Rubin platform:
- Trains AI for "a tenth of the token cost" and "a quarter as many GPUs" compared to Blackwell (10:53).
- Surprised industry by accelerating the roadmap—hardware already sold out for all of 2026.
- Nvidia evolving from 'chip company' to the “infrastructure layer of AI...far beyond just chips” (12:20).
-
Open Weight vs. Open Source
- Jeff: "Let’s not call it open source models, let's call them open weights." (44:02)
- Open models favored for broad adoption, but distinction is significant.
3. Robots & ‘Physical AI’: Reality and Limitations
-
Humanoid and Utility Robots
- Shift from "unrealistic showpieces" to toys for kids and specialized helpers (laundry folders, robotic arms, etc.) (15:49, 17:26).
- "Physical AI" is the buzzword: From robot pets to industrial arms.
- "The laundry folding robot is like the holy grail, right?" – Leo Laporte (18:43)
- Yet, most shown are early concepts, not shipping; human dexterity remains hard to replicate.
-
AI for Kids: Promise and Peril
- Big push on 'emotional support' robots—examples include a “sock with an eye” (21:32), and a panda robot (“biometric affective AI Panda”) that interacts with children.
- Questions about privacy, manipulation, emotional development:
- “One person’s emotional support robot is another’s emotionally manipulative device...But it was in a sense, addressing a little bit of loneliness of that kind of thing.” — Jason Hiner (21:36)
- Audience poll from Jason’s newsletter: "Would you give a robot...to a child in your life? 50% said no right off the bat." (26:54)
4. Consumer vs. Enterprise AI: Where’s the Real Value?
- Consumer Products: Gimmicks and Gaps
- AI in appliances (smart fridges suggesting recipes, etc.) failed to impress the hosts—solutions in search of a problem (27:34).
- “Why would I want to spend that much more on a refrigerator?...all it does is look in the refrigerator and tell me…” – Jason Hiner (27:34)
- Enterprise Momentum
- Enterprise pavilion at CES is larger than ever (29:14).
- True AI momentum is with businesses, not individual consumers:
- Automation, domain-specific models, cost savings, productivity.
- Countertrend for 2026: Companies moving to smaller, cheaper, domain-specific models for better ROI (30:59).
- Developers are an important offshoot: “This is really where you’re first seeing what I would say is almost AGI.” – Leo Laporte (31:27)
5. AI Controversy, Ethics, and Policy
- X.AI’s Grok Model & Content Issues
- Controversy over generating deepfaked, nonconsensual or explicit images using AI ("non-consensual bikini pictures and nude pictures," 45:13).
- X.AI’s statement: “...if you can’t handle innovation, maybe log off.” (45:51)
- Despite backlash, massive new funding round for X.AI ($20 billion) to build supercomputers (46:31).
- Florida’s Anti-AI Push
- Gov. DeSantis pushes for "rejecting AI with every fiber of our being"—citing deepfakes, therapy bots, parental controls, data center restrictions, and more (61:20).
- Panel sees politics, not technology, at play: “I am shocked there’s politics in my politics.” – Paris Martineau (66:00)
6. Real-World AI: From Healthcare to Coding Assistants
-
OpenAI Launches ChatGPT Health
- ChatGPT Health—AI to assist with health records and medical questions (50:47).
- Caution emphasized: “...designed to support, not replace medical care. It is not intended…to diagnose or treat.” (51:10)
- 200M users already ask health questions of ChatGPT (53:24).
-
Coding, Vibe Coding & AI as Developer Platform
- Cloud Code/Claude—Anthropic’s AI—outshines Google’s team in rapid prototyping (98:30).
- “I’ve been very impressed with what you can do to cut through the noise.” – Leo Laporte (98:46)
- AI is transforming coding as an iterative, collaborative process—less about writing code, more about “producing” and sculpting software by command and review (100:27 onward).
-
Live Demo: ‘Tomato Toss’ Game Vibe-Coded by AI
- The crew ideates and builds a “throw tomatoes at Leo” game in real time with Claude—prompt, adjust, and deploy iteratively, no formal programming needed (102:22–134:45).
- “That’s pretty impressive actually, that you did that in about five minutes.” — Leo Laporte to Paris Martineau (134:01)
7. Reflections & AI’s Place in Society
-
Are We Using the Right Language for AI?
- The panel challenges “AI slop” label and anthropomorphizing—advocating a view of AI as a tool, not a thinking being.
- “We need to stop expecting of it superhuman capabilities. It is a tool.” – Leo Laporte (113:51)
- Historian Thomas Haigh cited for the idea that “AI” is really branding, and that companies may soon rebrand away from the term due to negative public perception (115:09).
-
Learning AI: No Textbook Approach
- The consensus: No one-size-fits-all course or book; best to learn by example, via open courses (Stanford, Harvard, MIT), YouTubers like Nate B. Jones, and hands-on experimentation (119:47-124:10).
Notable Quotes & Memorable Moments
-
“One person’s emotional support robot is another’s emotionally manipulative device.”
— Jason Hiner (21:36) -
“The laundry folding robot is like the holy grail, right?...these things are getting smarter and smarter.”
— Leo Laporte (18:43) -
“Consumers are, are just not really sure and they have some fear and some trust issues and rightly so. I think where we’re seeing most of the momentum in AI is actually businesses.”
— Jason Hiner (29:13) -
“A slingshot style tomato game with a moving target. That’ll be fun. Let me build this for you.”
— Claude, as prompted by Leo Laporte (106:07) -
“I think we need a new model of what AI is...It is a tool.”
— Leo Laporte (113:51) -
“Slop specifically refers to it being machine generated...It is derogatory. That’s correct.”
— Paris Martineau (114:54) -
“I’m overwhelmed. I can’t keep up. It’s happening too fast. All of this coding stuff is going too fast for me.”
— Quoting Andrej Karpathy (85:00)
Timestamps for Major Segments
- 00:50–04:30: Say hello to 2026; CES is bigger, busy, and "all about AI."
- 06:59–14:47: Nvidia keynote and Vera Rubin platform; Nvidia’s new identity as infrastructure layer, not a chip company.
- 15:49–27:34: Robots, physical AI, and the paradox: toys, “emotional support robots,” and the privacy/ethics question for products aimed at children.
- 27:35–32:28: Consumer products vs. enterprise focus; what’s actually useful at CES.
- 44:01–46:10: X.AI Grok controversy; open source vs. open weight debates.
- 50:47–56:01: ChatGPT Health and AI’s role in health/medicine; panel’s experiences and warnings.
- 98:30–134:45: AI coding assistants; panel builds a “Tomato Toss” game live, reflecting on how the process changes coding.
- 113:51–116:09: Are we talking about AI the wrong way? Branding, expectations, and public perceptions.
- 119:47–124:10: How do you actually learn AI? No textbook, try, experiment, and use open courseware.
- Catalogued throughout: Live anecdotes, memorable quotes, recurring humor, and spirited panel banter.
Additional Highlights
- Debate over whether “AI slop” is the right critique – is all bad output slop, or only machine-generated “meaningless” content?
- Growing list of “AI wearables”—memory pins, audio pins, etc.—the B and Humane pins, and skepticism they’ll catch on.
- Florida’s AI policy and general public AI skepticism as political theater more than technology critique.
- Practical advice: Use LLMs locally (Anything LLM, etc.), try open-weight models (Mistral, DeepSeek, etc.), and steer clear of the snake-oil side of the AI industry.
- Pick of the Week: Floor 796—an ever-growing, pixel art GIF loaded with layers of pop-culture and video game references.
Takeaway
As AI moves beyond buzzword status at CES and permeates industry as much as consumer gadgets, the episode delivers a sober but playful survey of what’s real and what’s illusion in the intelligent machines revolution. Whether it’s the persistent quest for the laundry-folding robot, the endless stream of emotion-robot toys, or live-coded tomato-toss games, AI’s present and future is less about HAL 9000 and more about a thousand practical, incremental tools—if you can spot the slop amidst the hype.
Panel final thoughts:
“Play with it. Try it. Just don’t expect superhuman magic.”
“If you want the AI game, let me know—expect to receive Leo-Tomato-Toss in your inbox soon.”
“We need to think differently about what we ask of these machines, and what we want from ourselves.”