Results
30 articlesSort
20:55ENJesse Hall: Build Voice Agents for Conversations, Not LeaderboardsJesse Hall · LiveKitLiveKit’s Jesse Hall explains how to evaluate voice agents as real-time systems: understand callers, handle interruptions, keep talking while tools run, and measure the delay people actually hear.
1:04:35ENCliff Weitzman on Owning GPUs, Voice AI and Building SpeechifyGuest Cliff Weitzman · Host Harry StebbingsSpeechify's CEO makes the case for owning part of a compute fleet, reflects on a missed business-software opportunity and discusses AI-assisted teams.
34:30ENNetic's Agents Answer the Phone When the Heat BreaksHost Sarah Guo, Elad GilFounder Melisa Tokmak tells Elad Gil how a two-year-old platform became the AI operations layer for HVAC, roofing, pet care, and other essential services.
15:12ENTwelve Open-Source AI Projects for Agents, Video, Coding and VoiceMatthew Berman · Forward FutureMatthew Berman surveys tools ranging from long-running agent frameworks to document OCR, skill security and local speech generation.
1:23:53ENCreators Cheer Filler-Word Removal and Boo Generative VideoHost Nathan LabenzDescript CEO Laura Burkhauser maps the same customers' three-speed reaction to AI: deterministic buttons they love, an agentic co-editor they want but scold, and generative video they actively hate.
1:37:50ENGranola Added More Customers Than Almost Anyone by Being Surprisingly UnambitiousHost Nathan LabenzCo-founder Sam Stephenson tells Nathan Labenz how a calm notes app for frazzled meeting days, one Slack-sharing loop, and a refusal to promise the burning-fire work built a $1.5 billion company.
41:38ENElevenLabs Hit $300 Million ARR by Building Voice Before Anyone Believed It Was the InterfaceHost Sarah Guo, Elad GilIn Poland, watching a foreign film is a special kind of pain: one narrator, every character, a flat delivery from start to finish.
44:47ENThe Last Mile Is Where Contact-Center AI Gets RealCresta CEO Ping Wu and Sequoia's Doug Leone on why 17 to 20 million agents, legacy systems, and last-mile integration will decide who wins customer experience.
23:39ZH → EnglishLi Mu: Adding Ten Million Hours of Speech to a Text LLM, Open-Sourcing Higgs Audio李沐 · Boson AILi Mu explains how the team added a huge amount of speech data to text LLM training, built an AudioLLM that can listen and speak, and open-sourced Higgs Audio. He covers architecture trade-offs, token compression, where the data came from, and how to run it on a 3090.
59:38ENElevenLabs Was Supposed to Be RoadkillCo-founder Mati Staniszewski on dirty audio data, prosumer virality, and why voice should become the default interface for technology.
38:43ENTwo in Five Doctors Want Out. An On-Call Cardiologist Built the Note That Lets Them StayHost Sarah Guo, Elad GilAbridge CEO Shiv Rao on treating the doctor-patient conversation as healthcare's source of truth, and why the company ran straight at the hardest health systems.
1:05:11ENThe Fake Five-Second Hold: Voice AI Is Already Negotiating With TruckersHost Nathan Labenza16z partners Olivia Moore and Anish Acharya walk Nathan Labenz through what is working in voice AI — call centers, coaches, companions — and what still is not.
30:09ENBilt Saved 65 Support Seats. Decagon's Founder Says Most Agent Ideas Still Won't SellHost Sarah Guo, Elad GilJesse Zhang on why customer support is the golden agent use case, why instruction following beats o1-style reasoning, and the two filters that kill most agent startups.
56:10ENDeepL Beat Google at Translation by Refusing to Be a General ModelHost Jacob EffronOnly 18 percent of the world speaks English—and a Cologne startup turned that fact into a $2 billion language-AI company.
32:07ENA 10-Person Team, One Click, and Two Years of Fake DeadlinesNotebookLM’s product and design leads explain how a source-grounded research tool became a viral two-host podcast generator—and what they refuse to add next.
34:09ENFive Independent Teams Found the Same 10-to-1 Mix That Beats TransformersHost Sarah Guo, Elad GilThe authors of S4 and Mamba argue that mostly state-space layers plus a dash of attention is the architecture that actually works—and they are shipping it first as real-time speech.
1:45:11ENAI already beats doctors at empathy. Catastrophe is still not the defaultHost Nathan LabenzNathan Labenz tells Sarah Hastings Woodhouse why he is an adoption accelerationist and a hyperscaling pauser, and lists eight reasons he will not join the hardcore doomers.
30:55ENGPT-4o's Early Demos: Voice, Tutoring, Meetings and AccessibilityMatthew Berman, Greg Brockman, Sal Khan, Imran · Forward FutureMatthew Berman examines OpenAI's May 2024 demonstrations and considers what real-time multimodal interaction might enable.
1:22:43ENFive billion minutes of calls: why Dialpad trained its own business LLMHost Nathan LabenzDan O’Connell on real-time speech, verticalizing the stack, and why he and Nathan Labenz disagree about when AI employees show up.
1:17:20ENHume wants well-being in the objective function, measured 53 emotions at a timeHost Nathan LabenzAlan Cowen tells Nathan Labenz how Hume AI reads face, voice, and language, sells custom models at $1.66 an hour, and wants long-horizon well-being to override every other AI goal.
1:20:55ENA Custom Voice Used to Cost $10,000. Ten Minutes of Audio Now Clones It.Host Nathan LabenzPlay.ht CEO Mahmoud Felfel tells Nathan Labenz how text-to-speech jumped from robotic cloud APIs to director-level voice generation — and why the company is racing to put brakes on abuse.
1:28:46ENThe Speech Company That Gets Paid to Label Its Own Training DataGuest Dan Kokotov · Host Lex FridmanRev VP of engineering Dan Kokotov tells Lex Fridman why a $1.25-a-minute human transcript, a 14% machine word-error rate, and a two-sided freelancer market are the same flywheel.
1:45:58ENNine of ten colleagues said far-field speech was impossibleGuest Rohit Prasad · Host Lex FridmanRohit Prasad, Amazon’s vice president and head scientist for Alexa, talks with Lex Fridman about conversational AI, the Alexa Prize, trust, and why a 20-minute chat is still years away.
1:09:31ENThe Staircase to an AI Tutor: Connor Wick on Speak, Moats, and Voice-First LearningHost Jacob EffronHow Speak's CEO is building a fluency product one model-capability step at a time, from specialized speech models to evals, pricing, and the future of education.
Load moreShowing 24 of 30








