generative experiences co. @generativexp

We may have found three new planets. NASA has a space telescope called TESS. It photographs the sky every few minutes and puts the pictures online for free. When a planet passes in front of its star, the star dims a little for a few hours. Find that dip repeating like clockwork around a distant star, and you have found a planet candidate. We built a pipeline with Claude that checks stars for those dips, and pointed it at 38,620 small, dim red stars that the big research groups mostly skip. Three passed every test. Each one dims on schedule, every time TESS has looked, from 2018 to now. Two are about the size of Neptune. One is the size of Jupiter, orbiting a star half the size of our Sun, 480 light-years away. Textbooks say that should almost never happen. They are candidates, not confirmed planets. A bigger telescope has to conduct further analysis. The write-up is under review at the American Astronomical Society, and NASA’s follow-up site has confirmed the path to list them for other astronomers. Everything is public, including a log of the 25 things we got wrong along the way. Thanks to @NASAGoddard (TESS), @MIT, @SpaceTelescope (MAST), @caltechipac (ExoFOP), and the @esa. This research was made possible because of @AnthropicAI (Claude) and @modal (compute). The candidates, the figures and the code: genexperiences.com/tess-hunt

Talked to multiple founders yesterday who have been building their products over the last 3 years. Started manual code. They feel behind and slower than companies that started this year with all code written by agents. I have never seen such an order of magnitude difference.

@emmettshine amazing!!

fable 5.2 is amazing ngl

@axhiii88547 im team this x.com/generativexp/s…

generative experiences co. @generativexp

3 weeks ago

We're sharing a preview of an internal model we're code naming "Haydn-01", unlike traditional text-to-speech, or audio understanding modes like @OpenAI ChatGPT voice, @Google Gemini TTS, or @ElevenLabs, etc. Haydn-01 does 3 things that no other model does in a single

1 0 1 161 1

@gmkurtzer x.com/generativexp/s…

generative experiences co. @generativexp

3 weeks ago

We're sharing a preview of an internal model we're code naming "Haydn-01", unlike traditional text-to-speech, or audio understanding modes like @OpenAI ChatGPT voice, @Google Gemini TTS, or @ElevenLabs, etc. Haydn-01 does 3 things that no other model does in a single

1 0 1 161 1

@adamho modern music for sure x.com/generativexp/s…

generative experiences co. @generativexp

3 weeks ago

We're sharing a preview of an internal model we're code naming "Haydn-01", unlike traditional text-to-speech, or audio understanding modes like @OpenAI ChatGPT voice, @Google Gemini TTS, or @ElevenLabs, etc. Haydn-01 does 3 things that no other model does in a single

1 0 1 161 1

@TheGeorgePu x.com/generativexp/s…

generative experiences co. @generativexp

3 weeks ago

We're sharing a preview of an internal model we're code naming "Haydn-01", unlike traditional text-to-speech, or audio understanding modes like @OpenAI ChatGPT voice, @Google Gemini TTS, or @ElevenLabs, etc. Haydn-01 does 3 things that no other model does in a single

1 0 1 161 1

@danellisona a scribe once we're done with digital bytes, we're going to need to source life experiences to train models on from somewhere. start with 10 and just go door knock and you'll be suprised how many people just start talking about their life.

@solopribuilds ask them to pick an industry they like and make them think about how it can flourish with ai

@OpenAI @Google @ElevenLabs linkedin saw it first

We're sharing a preview of an internal model we're code naming "Haydn-01", unlike traditional text-to-speech, or audio understanding modes like @OpenAI ChatGPT voice, @Google Gemini TTS, or @ElevenLabs, etc. Haydn-01 does 3 things that no other model does in a single architecture: 1) Audio classification (pass in a speech sample and ask it if the person seems angry, sad, happy...) 2) Real-time conversation (TTFC @ 2.1s / 0.88 RTF) 3) Non-speech audio understanding (identifies animal noises, crowded places, loud environments, and refuses to predict understanding when not possible) Thanks to open contributions from @mmastrac (djev), @trynarilabs (DiaTTS), @GoogleDeepMind (DiffusionGemma), @typesafeai (Jev), vLLM, @modal (compute), and Hugging Face credits: TaDiCodec (Amphion, NeurIPS '25) and Future (artist), Epic Records, Freebandz

About a year ago, the Magenta team at @GoogleDeepMind worked with artists @hollyherndon and @matdryhurst to finetune Magenta RT, their open-weights real-time music model, on Holly's vocals. The base model was trained on roughly 190k hours of mostly instrumental stock music, so voice was one of its weakest areas out of the box. Post-training on Holly's vocals turned that weakness into a specialization. The general musical ability from pre-training is still there, but the output distribution has shifted toward sounding like Holly. For any organization sitting on specialized knowledge, post-training your own model is the next step.

Jesse Engel @jesseengel

a year ago

We're thrilled to have worked with Holly Herndon to train a model finetuned on her vocal style to share with all of you. Playing with Holly’s model, you can use additional controls to get vocalizations that contour the music or "color" the instruments with a timbre that

3 6 42 12K 12

@anitakirkovska is the number of people who associate the word transformer to ai vs. the cartoon going up or down should be the index with track. the statistics on understanding of ai could be really cool.

High speed rail would make it the best coast in the world.

@shirshakchavan linkedin

fundamentally this lacks taste

Tavus @tavus

2 months ago

Introducing Phoenix-4.5, the fastest and most expressive real-time human rendering model on the market. More natural than ever before: richer facial animation, more expressive emotion and micro-movements, and movement that now extends through the upper body. It is the closest

99 89 690 198K 532

We're experimenting with music. Introducing J16, an infinite playback engine that is laying the foundation for new music experiences. Check out the demo. This is not a public model. This is an experiment and access is limited to select industry partners and artists we hope to collaborate with. ie. @hollyherndon

@kickingkeys another certified banger

if charlie damelio made jeans for men would you buy it?