Jun 22, 2026 · 32m · big-technology
Are AI Glasses Over? + Big Technology AI Summit Audience Questions
Alex Kantrowitz · 10m spoken Ranjan Roy · 10m spoken Sarbjeet Johal · 45s spoken Maria Serra Roberts · 37s spoken
0:00 / 0:00
▶ Watch on YouTube →
gold bands on the timeline = statements, start to end. Hover to read, click to jump. CC turns on captions
Recorded live at the Big Technology AI Summit, hosts Alex Kantrowitz and Ranjan Roy debate the viability of AR smart glasses before engaging in an interactive Q&A on autonomous agents, enterprise architectures, biosurveillance, and tech ethics.
How this conversation actually went
Every chapter scored 0–10 on four independent dynamics. Hover any point for the reasoning behind the score. Alex holds 38.9% of the talking time here. How this is scored →
Alex as informed peer 6.1 Guest teaching 4.6 Guest disagreement 3.3 Alex pushing back 3.6
speaking balance: gold is Alex, purple is the guest (3 minute bins)
Sharpest disagreement ▶ 16:40 Ranjan cuts off Alex's travel agent example
Ranjan aggressively interrupts Alex's goal-mode example to mock the AI industry's fixation on flight and travel booking use cases.
Hardest push from Alex ▶ 3:35 Alex rejects the premise of face computingAlex forcefully pushes back against Ranjan's enthusiasm for AR glasses, declaring that upgraded smartphone assistants will render face-worn computing unnecessary.
Biggest teaching moment ▶ 23:30 Ranjan explains agent context architectureRanjan educates the audience and host on the technical limitations of dumping huge context windows into agents versus architecting structured data chunking.
Alex holds their own ▶ 19:10 Alex analyzes AI advertising mechanicsAlex demonstrates strong domain knowledge by breaking down how AI companies will transition from high-touch brand ads to direct-response monetisation.
the scores for every segment, with the reasoning behind each
| Chapter | Topic | Alex as informed peer | Guest teaching | Guest disagreement | Alex pushing back | Why |
|---|---|---|---|---|---|---|
| Debating Snapchat Spectacles and the Future of AR Glasses | 7 | 4 | 5 | 7 | Alex makes a strong case that dedicated AR face hardware is doomed, arguing the smartphone and improved Siri will dominate AI interactions. Ranjan pushes back in defense of AR form factors, creating a lively debate around hardware utility and market viability. | |
| Audience Q&A: Planning Horizons and Evolving Voice Interfaces | 6 | 3 | 2 | 2 | Audience member Sarbjeet asks about company planning horizons. Ranjan and Alex collaborate to discuss evolving voice and dictation interfaces as well as the rapid pace of frontier model capabilities. | |
| Audience Q&A: AR Hardware Architectures and Everyday Use Cases | 6 | 4 | 3 | 4 | Audience member Kyle asks about AR architectures. Alex quickly pushes back against external compute pucks, while Ranjan suggests offloading processing to the iPhone in one's pocket. | |
| Audience Q&A: Autonomous Agent Benchmarks vs. Real-World Execution | 6 | 5 | 5 | 4 | Sasha asks about agent benchmarks. Ranjan dismisses benchmark hype by pointing out the gap between lab benchmarks and real enterprise data, teasing Alex over using travel examples for agent use cases. | |
| Audience Q&A: AI User Profiling and Advertising Monetization | 7 | 4 | 3 | 3 | Addressing an audience question on AI advertising, Alex details the shift from brand ads to direct response while highlighting the privacy risks of LLM profiling, with Ranjan agreeing on the scale of user intimacy. | |
| Audience Q&A: Agent Context Drift and Information Architecture | 5 | 7 | 2 | 1 | Saoji asks about agent drift and learning layers. Alex cedes the floor to Ranjan, who provides deep enterprise context on context-window management and data chunking architecture. | |
| Audience Q&A: AI in Biosurveillance and Biosecurity Safeguards | 6 | 5 | 4 | 6 | Chris asks about AI biosurveillance. Alex turns the question around to probe whether frontier biology capabilities should be gated, prompting a discussion on dual-use pathogen risk versus open detection. | |
| Audience Q&A: US-China AI Dynamics and Frontier Model Selection | 6 | 5 | 2 | 2 | Heidi asks about US-China AI relations. Alex expresses cautious optimism for health collaboration while Ranjan highlights the sudden industry-wide adoption of DeepSeek models due to inference costs. |
Statements from this episode (14)
Kantrowitz: Face-worn computing fails; the iPhone is the true AI device
“And so maybe this idea of we have to wear the computing on our face is something that like kind of sounds good in concept, but model after model, it's not. And the AI device is the iPhone.”
Roy: Jony Ive and OpenAI's hardware collaboration may use a pin form-factor
“I mean, we humane tried with the pin. I still think maybe some kind of pin is going to be around Johnny Ives pin at open AI.”
Kantrowitz: Wearable devices requiring an external compute puck will fail
“Anything, any device that requires a puck not working.”
Roy: Apple AR glasses could succeed by offloading heavy compute to iPhones
“It makes me think Apple still has a good chance in this space because the iPhone can do all of the heavy lifting versus this thing on that is really heavy on your head.”
Kantrowitz: Apple exclusively marketed the Vision Pro as an isolated, solo device
“Apple never advertised that as a social device. All the marketing was just You're sitting alone at home, and all you're doing is the Vision Pro. Nobody else exists.”
Kantrowitz: AI autonomous coding endurance jumped from 30 minutes to 18 hours
“You saw these models, they could not code autonomously for more than 30 minutes in 20, 23. And now they can code autonomously for the human equivalent of 18 hours, right?”
Roy: A massive gap remains between AI benchmarks and real-world enterprise execution
“I still think there's such a massive gap in terms of what does the data look like? What is the actual problem? Does the customer, Or the person actually understand the goal in a clear enough way that they're able to define it to kind of push that loop forward.…”
Kantrowitz: AI platforms will inevitably use personal interaction data for targeted ads
“So will AIs build an amazing sort of profile of you based off of all the personal data and then use that for ads? Most certainly.”
Kantrowitz: AI companies will aggressively pivot from brand advertising to direct response
“And obviously they always come out with the high touch brand and then they will like get you on the direct response, like super targeted stuff. Once they realize they can't make money on brand.”
Roy: Google Gemini will monetize via ads, but Anthropic will avoid them
“At least anthropic is not, I mean, I don't think they're going in that direction at all. Google and Gemini, Obviously, 100% will”
Roy: Jamming raw data into large context windows fails at agentic scale
“Six to 12 months ago, the approach was just jam as much information into whatever system you can. And we were kind of promised that it would just work and AI is just gonna, and like, there's, you know, you'd get this feeling that, oh wait, a hundred page PDF. …”
Roy: Mastering AI agent information flow will be a massive career opportunity
“I actually think this is going to be one of the most interesting, like professional opportunities and areas to be an expert in going forward is like, Being able to kind of, it's funny, like it is as much art as science right now, but I think like the more you …”
Roy: Due to cost, DeepSeek is discussed in every AI agent conversation
“The conversation around moving towards deep seek and adding it into your agentic process or adding Chinese models did not exist in any conversation I was in 12 months ago and now is in almost every conversation, at least as an option, because cost has become s…”