Voice and Audio AI
Speech recognition, voice agents, text-to-speech, audio generation, dubbing, call automation, and voice-first AI interfaces.
Hackathon Caps Models at 32B Parameters to Reward Tinkerable AI Apps
Build Small is a Hugging Face and Gradio hackathon organized around a hard constraint: every model used must be under 32 billion parameters. Yuvraj Sharma framed the rule as a way to move AI building away from dependence on giant hosted models and back toward systems that participants can inspect, fine-tune, run locally, and ship as working Gradio Spaces. Sponsor presentations from Black Forest Labs, OpenBMB, OpenAI, NVIDIA, Modal, JetBrains, and Cohere largely reinforced that premise, offering small models, credits, tools, and prize categories meant to turn the constraint into runnable projects rather than demos in name only.
Voice AI Benchmarks Understate Errors in Real Multi-Speaker Audio
Hervé Bredin of pyannoteAI argues that voice AI benchmarks often make speech-to-text look more solved than it is by evaluating cleaner, more single-speaker-like audio. In his talk, he shows Nvidia Parakeet scoring 11.4% word error rate on AMI meeting audio in the Open ASR Leaderboard but 26% in pyannoteAI’s run on the same dataset using the table microphone rather than headset audio. Bredin’s broader case is that conversational AI needs fine-grained speaker diarization and speaker-attributed transcription, because words alone do not capture who spoke, when they overlapped, or how real multi-speaker conversations are structured.
AI Voice Agents Are Beating the Average Customer-Service Rep
Tom Chen, chief product officer at Aircall, argues that AI voice agents should be judged against the average customer-service interaction, not the best human rep. In his account, the technology is already good enough for many routine calls, can handle far more concurrency at lower cost, and may improve satisfaction when customers are given a clear choice between faster AI service and a human agent. The main constraint, Chen says, is often not the model but the undocumented company knowledge the agent needs to resolve issues.
Coding Agents Are Becoming a Managed Workforce Inside Conductor
Conductor CEO and co-founder Charlie Holtz argues that AI coding tools should be managed more like a team of workers than used as autocomplete inside an IDE. In a demo of how he uses Conductor to build Conductor, Holtz shows a workflow built around starting multiple agent workspaces, reviewing their pull requests, and merging only the work that passes human judgment. He says the shift makes prompts, architecture, review discipline, and “slop-free” parts of the codebase more important as hand-written code becomes less central.
Useful AI Systems Are Emerging Inside Controlled Enterprise Workflows
TBPN’s latest discussion framed the commercial AI moment less as a race to looser autonomy than as a shift toward bounded systems. Across Microsoft’s Build announcements, Suno’s funding, creator films, stablecoins, crypto markets, cybersecurity, and workflow software, the central argument was that AI becomes useful when it is embedded in infrastructure that can price, route, audit, secure, or constrain it. John Coogan and guests applied that lens most directly to Microsoft’s agent strategy, where Azure and Microsoft 365, not a new phone, become the controlled operating environment for enterprise agents.
The Model Alone Is No Longer the AI Product
At AI Engineer Melbourne 2026’s Day 1 keynote program, speakers including Shawn Wang, George Cameron, Sarah Sachs, Igor Costa, Vamsi Ramakrishnan and Geoffrey Huntley argued that AI engineering has moved beyond picking the strongest model. Their shared case was that useful AI products now depend on the systems around models: harnesses, routing, evals, memory, state, latency budgets, deterministic tools and cost controls. The model still matters, but the keynote program framed product advantage as an architecture and economics problem, not a leaderboard problem.
Screen Fatigue Is Driving New Markets for Physical Consumer Products
Sam Parr and Shaan Puri use a My First Million episode to test seven unconventional business ideas against a narrower question: whether each points to real demand or just novelty. Their strongest cases are for anti-phone hardware, social wellness formats, physical screen-free media and VR trade training, where they argue odd-looking products attach to existing pressures such as phone addiction, screen fatigue and labor shortages. They are more skeptical of ideas that rely on unverifiable claims or inflated mission language, including AI pet translation and clinical-trial prediction markets.
Travelers Deploys AI Claims Assistant Nationwide After Eight-State Pilot
Travelers’ claims CIO Erik Roen argues that putting an AI assistant into first notice of loss required changing the operating model around claims, not just adding a model to a call flow. In a conversation with OpenAI chief revenue officer Denise Dresser, Roen says the insurer moved from an eight-state pilot to countrywide deployment by pairing OpenAI’s technology with cross-functional business ownership, continuous evaluations, near-real-time monitoring and fail-safes for a workflow that helps customers decide whether and how to file a claim.
Language Models Are Becoming the Bottleneck in Video Generation
Ethan He, who worked on NVIDIA’s Cosmos world model and xAI’s Grok Imagine, argues that the next major gains in video generation will come less from diffusion models alone than from language models, agents, and context management around them. In an interview with swyx and Vibhu Sapra, He describes Grok Imagine as a fast-built example of that shift: diffusion renders pixels, while language systems increasingly rewrite prompts, plan clips, call tools, manage memory, and turn short generations into longer, editable video.
A Two-Hour AI Prototype Let Museum Visitors Talk to Statues
Joe Reeve of ElevenLabs argues that his “talk to a statue” prototype mattered less as a museum product than as evidence of what can now be assembled quickly from existing AI APIs. Built in Cursor in about two hours, the app identifies a photographed statue, generates historical context and a plausible voice, spins up an ElevenLabs agent, and starts a conversation in roughly 30 seconds. Reeve says the harder remaining questions are institutional rather than purely technical: who authors the object’s story, what voice it should have, and how multimodal voice interfaces should work.
Sarvam and NVIDIA Build Full-Stack Sovereign AI Infrastructure for India
Sarvam co-founder Pratyush Kumar argues that India’s AI sovereignty cannot mean putting Indian-language interfaces on foreign-built systems. In a NVIDIA-backed account of Sarvam’s work, he describes a full-stack effort to build foundational models, data pipelines, inference systems and developer APIs inside India, using NVIDIA H100 clusters and NeMo tooling to process Indian-language data at scale. The case is that voice-first AI for India’s population requires domestic capability across data, models, applications and accelerated-compute expertise.
Voice Agents Need Colocated Models to Stay Under One Second
Rishabh Bhargava of Together AI argues that production voice agents are now constrained less by demos than by a sub-second engineering budget spanning speech-to-text, LLMs, text-to-speech, networking, and scaling. In his account, users notice delays above 500ms and abandon calls around one second, making even 75ms network hops material once model latency is optimized. The practical architecture remains a cascade, he says, because it lets teams control tool calling, evaluation, and reliability while speech-to-speech models still lag on production requirements.
Personal AI Systems Need Separate Layers for Memory and Autonomy
Nathan Labenz opens his personal AI infrastructure to a security audit by Daniel Miessler, showing a system that combines a high-context Claude Code “second brain” with lower-access autonomous agents for operational work. Their central argument is that useful personal AI should not collapse memory, authority, and autonomy into one assistant: raw personal history should be preserved and audited, while agents that act in the world need narrower permissions, clear roles, and containment. Miessler frames the longer-term model as an assistant that navigates from current state to ideal state while continually pruning obsolete scaffolding as models improve.
Hugging Face Ships a $299 Hackable Robot for Voice AI Experiments
Andres Marafioti argues that Hugging Face’s Reachy Mini is meant to move robotics experimentation out of expensive humanoid hardware and into a $299-to-$449 open-source platform that users can assemble, repair and modify themselves. The robot’s most-used application is conversation, and Marafioti’s account ties its social ambition to a technical stack built for low-latency speech: Parakeet transcription, Qwen 3.5 27B, and an optimized Qwen3 TTS implementation that he says improved from 0.8x to 5.8x real time.
ElevenLabs Music v2 Adds Section Editing and Mid-Track Genre Shifts
ElevenLabs’ launch walkthrough for Music v2 presents the model as a more controllable generative music system, not only a higher-quality one. Alec Wilcock says the new version improves vocals, instrumentation, arrangement, multilingual output and dense vocal delivery, while adding section-by-section composition, targeted inpainting and the ability for one song to move between genres without losing coherence. The company also says the model is trained on licensed data and that generated tracks are cleared for commercial use.
Claude Code Reverse Engineers Viking VoIP Phone’s Undocumented Configuration Protocol
Boris Starkov of ElevenLabs presents the Viking K-1900D-IP phone as a reverse-engineering case study in which Claude Code turned an unusable, undocumented VoIP handset into a working AI demo. Starkov argues that Claude did the investigative work: discovering a two-letter command protocol, brute-forcing valid registers, intercepting the manufacturer’s Windows XP-era software through a TCP proxy, and deriving the one-byte checksum needed to write persistent configuration. His account is also a claim about agency in hardware work: he says he acted largely as Claude’s hands while Claude orchestrated the protocol break.
ElevenLabs Says Dubbing v2 Preserves Performance Across 90 Languages
ElevenLabs is introducing Dubbing v2 alpha as an AI dubbing model built around preserving the original speaker’s performance, not just translating a transcript. The company says the system conditions directly on source audio so tone, pacing, emphasis and emotional delivery can carry across more than 90 languages, with sync-aware translation adapting phrasing to fit the timing of the original. ElevenLabs is positioning the launch for creators, marketers and studios that want automated localization without building a separate dubbing pipeline.
Voice Will Become the Default Interface for Enterprise AI
Luiz Domingos, chief technology officer of Mitel, argues that enterprise AI has moved past pilots and into communications workflows where latency, compliance, auditability and human oversight determine whether systems can be deployed. In a conversation with Craig Smith, Domingos says cloud-only AI will not meet the needs of real-time voice and regulated industries, and that edge and hybrid deployments will become central. His larger prediction is that enterprise AI will increasingly be accessed by voice rather than screens, especially for frontline workers whose jobs do not fit a desktop interface.
ElevenLabs Adds Licensed Stan Lee AI Voice to Creator Tools
ElevenLabs is introducing an approved AI replica of Stan Lee’s voice through a partnership with Stan Lee Universe, positioning the late comic-book creator as a licensed feature inside its voice and creator tools. The company says users can request to license Lee’s voice for projects, hear it in Eleven Reader, generate Stan Lee cameos, and use Stan-inspired music, while repeatedly framing the launch around official authorization, rights ownership, and Lee’s mythology of stories being carried forward.
ElevenLabs Launches Music v2 for Licensed Commercial AI Song Generation
ElevenLabs is presenting Music v2 as a licensed-data AI music model built to generate vocal-led tracks from detailed natural-language prompts, not just loops or backing beds. The launch materials argue that the model can produce finished-sounding, one-shot outputs across styles and languages, while adding workflow features such as targeted inpainting, section-by-section composition, and deployment through ElevenMusic, ElevenCreative, and a forthcoming ElevenAPI.
Synthetic Intimacy, Surveillance, and Stimulation Are Raising the Cost of Impulse
Chris Williamson’s inaugural Mostly Wise conversation with Andrew Huberman, Matt McCusker and Tom Segura uses health advice, comedy, AI replicas and conspiracy talk to examine where useful tools become distortions. Huberman repeatedly argues for moderation and mechanism over slogans — from low-dose tadalafil and sleep protocols to cannabis, sunscreen and self-control — while Segura and McCusker test those claims against comedy, parenting and lived experience. The broader case is that modern life increasingly requires judgment about thresholds: when optimization becomes rumination, evidence becomes pattern-seeking, and synthetic intimacy or surveillance starts to reshape ordinary behavior.
Google’s GenAI Stack Turns Multimodal Prompts Into Application Pipelines
Google DeepMind’s Paige Bailey and Guillaume Vernade argue that Google’s generative AI stack is being organized as an application pipeline rather than a set of isolated models. In a three-hour workshop, Bailey showed AI Studio turning multimodal Gemini prompts into inspectable API calls and generated apps with auth and Firestore, while Vernade used Gemini, Nano Banana, Veo and Lyria to illustrate, animate and score The Wind in the Willows. Their case is that builders can now orchestrate prompt, code, media generation and deployment in one workflow, even as the demos exposed seams that still require engineering discipline.
Divergent Says Software-Defined Factories Can Build Drones in 71 Days
Lukas Czinger, co-founder of Divergent Technologies, argues that the bottleneck in defense hardware is not design but the tooling and fixed production lines that make iteration slow once a product leaves prototype. In a livestream interview, he said Divergent’s software-defined factory can move autonomous aircraft and other complex systems from digital design into production without rebuilding the supply chain around each change, citing a 71-day clean-sheet build of a flyable small uncrewed aircraft as proof of the model.
AI’s Bottlenecks Shift From Model Demos to Compute, Rights, and Institutions
AI, in TBPN’s latest discussion, is no longer treated mainly as a product demo but as a question of infrastructure, financing and institutional adoption. The strongest evidence came from SpaceX’s AI-heavy IPO framing, Anthropic’s reported move toward operating profit, and OpenAI’s claimed Erdős breakthrough, which the speakers used to challenge the “AI is a scam” critique. The unresolved issue is not whether the technology matters, but how quickly compute capacity, rights regimes, regulation and existing institutions can absorb it.