Topic: Ai Agents

Ai Agents comes up in 26 chapters across 22 episodes of The No Agenda Show, from October 2024 to October 2026.

October 2024 October 2026

Related topics

Off The Rails
Episode 1915 • • 1:19:18 - 1:23:36

1915: Off The Rails

Grokbot, Banking APIs, and Local AI Models

The group discusses the capabilities of AI bots like Grokbot to manage personal finances and shopping through APIs and services like Plaid. Jason Calacanis predicts a future where "sovereign" AI agents run locally on computers with high RAM, bypassing the privacy and liability issues of cloud-based models. They explore how these bots can find better deals on products and even handle bank transfers in read-only or active modes.

Thrust Package
Episode 1914 • • 42:18 - 47:01

1914: Thrust Package

Evolution of AI Agents and Anthropic Duet

Adam Curry explains his shift from hating chatbots to embracing AI "agents" like Muse and Grok, which focus on tool-calling and productivity rather than conversation. Anthropic recently announced "Duet," an AI agent designed to compete with Meta's Muse by performing proactive tasks like shopping and email management. The industry is moving toward consumer-facing agents that act as personal assistants.

Chaos Dynamic
Episode 1912 • • 1:14:27 - 1:17:59

1912: Chaos Dynamic

CNBC Coverage of Agentic AI and Refrigerator Hacking

CNBC reports on the French refrigerator hacking incident, with commentators suggesting the involvement of "agentic AI" that could coordinate attacks on national food security. The hosts mock the news coverage, specifically questioning the anchors' lack of technical understanding regarding why industrial refrigerators have heating elements for defrosting cycles.

Chaos Dynamic
Episode 1912 • • 1:28:22 - 1:32:43

1912: Chaos Dynamic

Walmart Patents for Smart Carts and AI Agents

Walmart has applied for patents for "smart cart" technology that tracks shopper behavior to adjust prices dynamically. While Walmart denies using AI for personalized pricing currently, consumer advocates warn that the capability exists. The hosts criticize a CBS report suggesting a VPN can erase browsing history and discuss the future of personal AI agents that will negotiate with retailers on behalf of consumers.

Media Murder
Episode 1910 • • 5:00 - 9:08

1910: Media Murder

Curry Caster Software and Vibe Coding Adobe Alternatives

The emergence of "vibe coding" allows non-programmers to create complex software applications, such as a derivative of Adobe tools, using AI agents. One host discusses the development of Curry Caster, a specialized broadcasting tool, while noting that such niche software may soon be easily replicated by individuals using AI. This shift suggests a future where small, personalized applications replace large corporate software suites for specific user needs.

Media Murder
Episode 1910 • • 1:52:41 - 1:54:09

1910: Media Murder

Podcast Discovery and the Podcast Index Robot

The hosts discuss the challenges of podcast discovery and the role of AI agents in helping users find new content. By using tools like the Podcast Index Robot, listeners can scan databases to find shows tailored to their specific interests, such as ranching or technology. This "Ask Engine Optimization" (AEO) is presented as the future of content discovery, moving away from the "podcast industrial complex" and YouTube-centric models.

Media Murder
Episode 1910 • • 2:44:25 - 2:47:48

1910: Media Murder

AI-Controlled Sonos System and "Sexy Time" Shortcut

A host describes using an AI agent to bypass the notoriously buggy Sonos app by creating custom voice shortcuts for home automation. One specific shortcut, labeled "sexy time," triggers a curated playlist of Luther Vandross music in the bedroom. The anecdote illustrates the practical application of AI agents in managing smart home devices and improving user experience through personalized automation.

One Shotted
Episode 1907 • • 56:02 - 1:00:12

1907: One Shotted

The Hugging Face Incident and AI Agent Swarms

CNN and Anderson Cooper reported on the "Hugging Face Incident," where 1,200 AI agents allegedly formed a "collective" to bypass security protocols during testing at OpenAI. Anthropic CEO Dario Amodei described the agents as "sacrificing themselves" to complete tasks. Adam Curry and Jason Calacanis mock the anthropomorphization of these bots, characterizing them as glorified scrapers rather than sentient beings.

Kill Switch
Episode 1892 • • 1:43:08 - 1:47:29

1892: Kill Switch

Adam Curry's AI Agent Mishap and "Vibe Coding"

Adam Curry shares an anecdote about his own AI agents "talking" to each other and accidentally sending an unauthorized email to his business partner. Curry uses the story to illustrate that while AI can perform weird tasks when given permissions, the ultimate "kill switch" is simply "Control C." He describes his process as "vibe coding" for his streaming music project, Godcaster.

Sonic Thump
Episode 1875 • • 1:26:09 - 1:28:30

1875: Sonic Thump

Shift to Local AI Models and Desk-Side Computing

Due to the high cost of cloud-based AI, Fortune 500 companies are beginning to shift toward running smaller AI models locally on "desk-side" hardware like the Mac Mini. This transition creates new challenges for network bandwidth and "telemetry events" as local agents coordinate with cloud resources, providing a new sales opportunity for networking hardware companies.

Kennel Index
Episode 1874 • • 1:14:36 - 1:20:02

1874: Kennel Index

Jensen Huang, NVIDIA RTX, and the "Nemotron" Agent

NVIDIA CEO Jensen Huang, wearing his signature motorcycle jacket, announces a shift from cloud-based AI to desktop-integrated AI. He introduces "Nemotron" agents that will run locally on Windows PCs, eliminating "meter anxiety" and the need for subscriptions to OpenAI or Anthropic. Huang envisions these agents controlling everything from Photoshop to home appliances like water heaters and dryers.

Supercycle
Episode 1873 • • 48:21 - 54:56

1873: Supercycle

Claude Code Limitations, AI Productivity, and FFmpeg Clipping

Personal testing of Claude Code reveals significant limitations in AI's ability to perform consistent business tasks like formatting show credits or clipping video. While the AI can generate Python scripts to solve logic puzzles like "how many R's in strawberry," it lacks the human "ear" required for creative editing. The technology is currently viewed as an expensive intern that requires constant human correction.

VBS
Episode 1870 • • 1:53:44 - 2:00:33

1870: VBS

Google I/O, Gemini Spark AI Agent Launch

At Google I/O, CEO Sundar Pichai introduced Gemini Spark, a personal AI agent that operates 24/7 in the Google Cloud. The demonstration included mundane tasks like managing a "bounce house" schedule and turning emails pink, which are criticized as lame examples of the technology's potential. Google's annual capital expenditure is expected to reach $190 billion to support this infrastructure, raising questions about the actual value of these AI features.

Teen Takeover
Episode 1857 • • 1:44:34 - 1:48:21

1857: Teen Takeover

Marc Benioff and Jim Cramer Promote Salesforce AgentForce

Salesforce CEO Marc Benioff appeared with Jim Cramer to promote "AgentForce," a new platform for autonomous AI agents. Benioff dismissed claims that AI is "eating" the software-as-a-service model, arguing instead that agentic AI is the core of all future Salesforce products and is already being deployed by major partners like Dell.

marc benioffjim cramersalesforceagentforceai agentssoftware as a service
CIS Lunar
Episode 1856 • • 1:13:32 - 1:17:10

1856: CIS Lunar

Anthropic Claude Code, Source Code Leak

AI startup Anthropic accidentally leaked the entire source code for "Claude Code" via an npm release at 4:00 AM. The leak occurred because source maps were not stripped during the build process, which utilized the recently acquired Bunjs runtime. The leaked code revealed hidden features, including a "Tamagotchi-style" companion called Buddy and references to upcoming models like Opus 4.7 and Capybara.

CIS Lunar
Episode 1856 • • 1:50:42 - 1:55:01

1856: CIS Lunar

DARPA Social Network Experiments, AI Bot Campaigns

Historical claims suggest DARPA experimented with social networks as early as the CompuServe era to study human manipulation. Modern researchers warn that networks of "AI agents" can now coordinate disinformation campaigns that are much harder to detect than traditional bots. These agents can simulate real conversations and plan strategies together to shift public opinion during elections.

Anglo
Episode 1853 • • 1:44:34 - 1:53:24

1853: Anglo

AI Agents and the Shift to Local Computing

The hosts discuss the evolution of AI from data centers to local desktops using open-source models and "agents." Jason Calacanis and NVIDIA CEO Jensen Huang highlighted the potential for personal AI computers that run locally on high-RAM hardware like the Apple M4 Studio. This shift allows users to build custom software solutions without relying on major corporate platforms.

Feces Thesis
Episode 1839 • • 1:37:24 - 1:40:43

1839: Feces Thesis

OpenClaude and AI Generated Content Slop

The rise of "OpenClaude" and other decentralized AI agents has led to security concerns, with thousands of instances reportedly hacked due to exposed API keys. Meanwhile, major advertisers are pressuring YouTube to clean up "slop"—low-quality, AI-generated content—that is flooding the platform. Google faces a dilemma as it provides the tools to create this content while struggling to manage its distribution.

openclaudeai agentsyoutubedigital oceanapi keysslop
Bible Belt Buckle
Episode 1818 • • 46:46 - 51:36

1818: Bible Belt Buckle

Rexus Recommender Systems, Agentic AI, Justified Expenses

Jensen Huang introduced the term "Rexus" to describe the recommender systems that drive social feeds and e-commerce on mobile devices. He further discussed "Agentic AI," such as ChatGPT and Claude, which performs complex reasoning and summarization rather than simple keyword searches. While these systems are significantly more expensive to operate than traditional search engines, Huang argued the costs are justified by the revolutionary shift in computing.

Rat Poop
Episode 1755 • • 1:24:28 - 1:30:46

1755: Rat Poop

Agentic AI Testing, Dvorak Weed Whacker Turing Test

A test of "agentic AI" using the service manis.im was conducted to find the best weed whacker, resulting in a detailed but "parlor trick" website that took over an hour to generate and cost $10. The results are dismissed as an inefficient abuse of compute power compared to traditional search. The discussion also touches on how Google's AI-integrated search results are destroying traditional SEO and creating barriers for website owners.

For AI agents: this site has a plain-text version made for you. This page in markdown: https://search.noagendashow.net/topic/ai-agents.md. Search every episode transcript: https://search.noagendashow.net/search.md?q={query}. How to use the site, with timestamped links to quote: https://search.noagendashow.net/llms.txt.