Sonic splitter @AudioshakeAi. Xoogler. Author of The Big Disruption. Fiction in @nytimes @VICE @WIRED. Mostly noise, some signal.
San Francisco
Joined April 2008
- Tweets447
- Following1K
- Followers4.9K
- Likes10.2K
Pinned Tweet
Today we're releasing @AudioShakeAI Multi-Speaker 2.0: one recording of several people talking goes in, a clean labeled track per person comes out — including the moments they talk over each other.
This is cool! But also, I have questions for the brain people:
1) Does speaking aloud help us think? I feel like I don't know what I'm saying half the time till I'm hearing myself say it
2) Loss of feedback? Tied to the above, I feel like hearing myself say something aloud helps me refine it, and recognize when something doesn't sound right (writing is an even better version of this)
3) You lose emotional cues (but longer term, you could get that from other facial expressions, at least from ppl with expressive faces)
We’ve raised $1.3M from YC & Afore Capital so you never have to type or speak again.
Typing is slow. Talking into your computer is awkward. Subvocal fixes both
Our glasses give you the speed of speech without you making a sound or touching a keyboard.
It tracks your internal speech muscles with non-contact sensing and decodes silent, almost invisible mouthing into text at more than 200 words/min.
Pre-orders are now live at:
subvocal.co/preorder?utm_sou…
You probably don't--and probably shouldn't!--think about AI when you're watching a live sports match, but source separation comes in handy if you want to:
- boost the call of the ref on the field
- drop a player screaming profanities but keep the crowd
- get rid of bleed from one source into a mic
....and do it all in under 15ms (to not screw up lip sync). Not noise reduction, actual separation.
All kinds of cool things that seemed impossible just 6 months ago.
@Dell wrote up the story of how we got there ⤵️
dell.com/en-us/blog/more-tha…
the rime models are really good!
and nice to see an AI company with some personality :)
introducing prism, the world’s first speech-to-speech model you can’t tell from a human in conversation
prism has:
> 300ms response time, 4x faster than OpenAI GPT-Live
> outperforms every other s2s model on intelligence (with NO reasoning)
> the first s2s model customizable through reinforcement learning
we’re a team of 35 with $35M in funding, and we’re unlocking s2s model adoption faster than openai, google, and x.
the first 100 people to sign up here get 1 month free: rime.ai/prism?utm_campaign=p…
With all the rage bait on here about human-made music being dead, worth noting that one of the buzziest Grammy albums this year shows exactly why humans aren’t going anywhere.
Swamp Dogg is 84. He cut his first record at 12. On this one, he ends a 50-year feud with another artist by singing a duet with him.
The album is good. Maybe a machine could make it /could make anything one day. What’s much harder to automate is the human mess: the career, the grudge, the fact that he’s still rocking at 84.
We’re humans. We like human stories.
jessica powell retweeted
Someone made a site that tracks what people are talking about at public city council meetings across the US.
The only thing discussed more than data centers?
Pickleball courts.
everyone saying that dazzle/marissa article is bleak and proof you should "go direct," completely miss what is valuable about traditional media: you get in front of all the people you DON'T normally talk to. and i bet there are plenty of moms that read that same article and felt all the examples resonated, incl the one where she obviously meant its hard to encapsulate your child in a few words.
Spent last week with a bunch of NYC songwriters, and everyone seems a little bored with Suno and in love with @TrebloApp.
Turns out, if you build a machine that makes the "perfect" song, you delight the people who don't know how to make music.
But the people who love making music....They want ideas and inspiration, not someone to do it all for them.
Ever since I was a little girl, it was my dream to one day make a tutorial video
What does a Lite-Brite have to do with AI?
A finished mix is one picture made of many pegs. Source separation pulls every peg of one color (vocals, dialogue, effects) onto its own board. @themoko explains 👇
audioshake.ai/post/what-is-a…
What do a Lite-Brite and AI have in common?
Here’s source separation, explained with a children’s toy from 1967. 👇
a more in-depth (non-Lite Brite!) explainer on source separation: audioshake.ai/post/what-is-a…
Hey @basicfuntoys @Hasbro, ready to talk about that multi-million-dollar product sponsorship? 😂
I’m big with the burned-out white-collar middle-manager / plays-guitar-on-the-weekends demographic.