Audio & Speech

Today's Stories

1. 1

   Anthropic clashes with Trump's White House, rejects Washington playbook

   The Trump administration has twice taken actions against Anthropic—labeling it a "supply chain risk" in April after the company refused Pentagon contract language, and imposing export controls on its Mythos and Fable AI models two weeks ago following discovery of a jailbreak. OpenAI, by contrast, announced it was withholding release of GPT-5.6 at the U.S. government's request on the same day those controls were relaxed. Anthropic, valued at $965 billion（約150兆円） and preparing for an IPO expected in the coming months, has refused the flattery, donations, and appointment of Trump allies that other tech giants (Meta, Amazon, Apple) and OpenAI have used to stay in the administration's favor. Trump administration officials have publicly attacked CEO Dario Amodei as a "liar" with a "God-complex" and an "ideological lunatic," and accused the company of "regulatory capture." Continued hostility could make it harder to sell public market investors on the stock listing and significantly hamper the company's ability to develop advanced AI models.

   Anthropic CEO Amodei reportedly called Trump "a feudal warlord" in a now-deleted Facebook post, and his sister and cofounder Daniela Amodei donated to Kamala Harris's campaign. Unlike OpenAI's policy chief Chris Lehane and cofounder Greg Brockman (the largest donor to Trump Super PAC MAGA Inc.), Anthropic has made no similar hires of Trump-aligned figures to its leadership.

2. 2

   Netflix uses AI-generated Gene Wilder voice for Wonka reality show

   Netflix is premiering Wonka's The Golden Ticket on September 23rd, a reality competition based on the fictional Wonka universe. The show's voiceover uses an AI-generated version of Gene Wilder's voice, created in partnership with AI audio company ElevenLabs and with consent from Wilder's family. This extends Netflix's pattern of using AI-generated celebrity voices for content—the company has previously recreated voices of Michael Caine and Stan Lee. For viewers, it means encountering synthetic versions of iconic figures in new productions, blurring the line between archival and synthetic media in mainstream entertainment.

   The two-part finale airs on September 30th. The show will feature 12 golden ticket winners and their chosen partners competing in a high-stakes social experiment, with one champion crowned by the end.

3. 3

   Reddit ML community member struggles to replicate Pocket TTS model from research paper

   A developer attempting to implement Pocket TTS (a text-to-speech model from kyutai-labs) without access to official training code reported significant training failures. Despite training on single-speaker datasets (LJSpeech) and LibriSpeech, the model achieved low loss metrics but failed to generate meaningful speech at inference time, even on text from its training set. Attempts to fix the problem using scheduled sampling and noise injection did not resolve the issue. The absence of published training and fine-tuning code makes it difficult for researchers and developers to reproduce or build upon academic work. When models are released without implementation details, community members must reverse-engineer solutions, consuming time and resources that could be invested elsewhere.

   The developer is seeking guidance on what went wrong in their implementation effort. The core challenge appears to be a gap between training metrics (low loss values) and inference quality—a mismatch that suggests either a conceptual misunderstanding of the paper's approach or missing implementation details not covered in the published work.

4. 4

   Hacker News: What's the Best AI Voice Tool Beyond ElevenLabs?

   A user posted a question asking whether there are AI voice narration tools better than ElevenLabs, citing dissatisfaction with flat delivery and character inconsistency from that service. The user had also tried Seedance 2.0 (limited to 15-second samples) and paid $75 for human voice work that proved unusable. ElevenLabs remains a widely-used baseline for AI voice synthesis, but creators working on animation and character-driven projects are finding its emotional range and consistency inadequate for professional work. The gap between AI tools and paid human actors suggests demand for a middle ground—AI with better character depth and emotional tone.

   The thread highlights a real gap in current AI voice technology: no respondent appears to have named a clear alternative that solves the emotional tone and character consistency problems, indicating this remains an open product challenge in the voice AI space.

5. 5

   NagaTranslate: AI translation and voice system for India's low-resource Naga languages

   A developer has built NagaTranslate, a pipeline combining translation and speech synthesis for Nagamese, Ao, and Sema—low-resource languages spoken in Nagaland, India. The system uses a commercial LLM API for text translation (after initially trying a fine-tuned NLLB model) and includes speech synthesis via Whisper and VITS. Nagaland's native languages were historically oral with minimal written standards and parallel training data, making them difficult subjects for conventional machine learning. This project addresses a gap in AI coverage for underrepresented linguistic communities, demonstrating how commercial and open-source tools can be adapted for languages that lack large annotated datasets.

   The developer is sharing the architecture and soliciting feedback on how to improve the pipeline under strict resource constraints, indicating the project is still in development and may evolve based on community input.

6. 6

   Article lacks business news content

   The body is a Reddit post asking for recommendations on AI voice agent technology, not a news report of a business event. This is a user question seeking technical advice, not a news story suitable for business readers.

   No news event, timeline, or business development is reported in the body.