

275 episodes
by Craig S. Smith · English
Eye on A.I. is a biweekly podcast, hosted by longtime New York Times correspondent Craig S. Smith. In each episode, Craig will talk to people making a difference in artificial intelligence. The podcast aims to put incremental advances into a broader context and consider the global implications of the developing technology. AI is about to change your world, so pay attention.

AI transformation for large organizations; why “efficiency” isn’t the destination, how to measure value-added work, and what real transformation costs.
30 Jul 2026 · 1h 6m · 23 chapters
Synthesia’s CEO Victor Riparbelli argues AI video is shifting from one-way broadcasts to interactive, two-way “avatar” experiences—especially for K-12 and enterprise training—where students/employees…
28 Jul 2026 · 40m · 18 chapters
Defining life and building “synthetic cells” (spud cells) from purified molecules; using them as an engineerable platform for medicine and future bioeconomy, plus safety/biosecurity discussion…
21 Jul 2026 · 46m · 22 chapters
The episode discusses Virtana CEO Paul Appleby’s “AI factory reality check,” arguing that enterprises are investing in AI data centers faster than they can govern and operate them.
15 Jul 2026 · 45m · 23 chapters
Braden Rogers (Island) discusses securing AI browsers and agentic workflows when they move from consumer use into enterprise environments.
13 Jul 2026 · 56m · 22 chapters
Industrial AI “with a hard hat on”—what it takes to build production-grade AI for asset-heavy environments (factories, refineries, hangars, offshore sites) where failure isn’t an option, including…
10 Jul 2026 · 23m · 11 chapters
AI “agents” (models with tool/API access) create new security risk because data can move between agents and outside human visibility.
7 Jul 2026 · 48m · 26 chapters
BullFrog AI argues that Big Pharma’s ~50% Phase 3 failure rate can be reduced by using AI to (1) clean messy biomedical data, (2) find causal disease drivers and correct drug targets, and (3) rank…
5 Jul 2026 · 50m · 21 chapters
Why enterprise AI agents fail and why the “trust gap” is mostly caused by data issues (stale, incomplete, inconsistent semantics), not model intelligence; how Denodo’s data virtualization + universal…
2 Jul 2026 · 42m · 16 chapters
Philip Goff argues that modern science’s approach to consciousness began with a foundational mistake by Galileo: pushing consciousness (qualia like redness and the “what-it’s-like” character of…
29 Jun 2026 · 1h 1m · 22 chapters
Qure.ai’s Prashant Warier explains how AI is used in radiology and diagnostic workflows, including a fully autonomous TB screening system, and discusses evidence, deployment, and future “AI-first”…
20 Jun 2026 · 42m · 18 chapters
Genpact’s “Autonomy by Design” research on why only 12% of enterprises generate business value from AI, and what it takes to scale AI into production workflows—especially with agentic AI.
18 Jun 2026 · 59m · 27 chapters
India’s shift from AI talent exporter and IT services provider to “architect of the global AI order,” emphasizing responsible, inclusive AI and a “techno-legal” approach to trust.
16 Jun 2026 · 56m · 23 chapters
The episode argues that enterprise AI is shifting from “assistants” to agentic workforces orchestrated through Slack.
13 Jun 2026 · 53m · 27 chapters
Zendesk’s agentic AI for customer service, including its claim that AI can resolve 90% of tickets and improve over time via a “resolution learning loop.” Shashi Upadhyay argues service will shift…
12 Jun 2026 · 57m · 28 chapters
The “100,000 agent problem” in enterprises—when many AI agents act autonomously on behalf of employees, trust/governance becomes the bottleneck.
6 Jun 2026 · 1h · 26 chapters
Aircall’s Tom Chen discusses why AI voice agents are accelerating adoption, how they fit alongside human agents, and what Aircall’s “assistance technology” does for sales/support teams.
4 Jun 2026 · 57m · 15 chapters
Evolutionary AI and neuroevolution as the “future of AI” for creativity and surprising solutions—contrasted with today’s gradient-descent and reinforcement-learning approaches.
2 Jun 2026 · 1h 4m · 29 chapters
How LogicMark uses AI and connected sensing to help seniors age at home longer—especially by improving fall detection (reducing false positives) and adding predictive analytics for events like falls,…
1 Jun 2026 · 53m · 18 chapters
How enterprises should modernize to adopt AI in communications—shifting from AI “strategy decks” to operational, workflow-embedded AI—plus why voice is becoming the natural AI interface.
28 May 2026 · 55m · 25 chapters
Whether ChatGPT/LLMs are conscious or “understand” language; what’s missing for sentience; and nature-inspired directions for generative AI and future “agents.”
28 May 2026 · 57m · 17 chapters
How children’s “data profiles” can be created before birth through adtech and data brokers, and how Proton’s “Born Private” aims to start a privacy journey by anchoring families to a private email…
28 May 2026 · 56m · 26 chapters
Steffen Kruse (Macrocosmos) explains how to train large language models without centralized “billion-dollar” GPU data centers, using BitTensor’s blockchain as a coordination/reward layer and…
25 May 2026 · 47m · 16 chapters
EY’s Errol Gardner argues the biggest barrier to AI (especially agentic AI) adoption in enterprises is not the technology but people: leadership, sponsors, middle managers, workers, and workforce…
22 May 2026 · 55m · 22 chapters