180 karma · joined August 11, 2013
Oh okay, it links to https://techcrunch.com/2018/11/28/google-dragonfly-letter, which says that a source within Google provided it to them, and they print it "in full", but sans signatures.
The more I think about these things, the more I distrust cloud providers, and want my own hardware.
Do you really trust these companies enough to hand them the keys to all your data? Is there really any way to provide secrets to your app without trusting the hosting provider?
Blue Dream
Green Crack
GG4
Dutch Treat
CinexBut anyway, I am hoping to make this role of teaching and babysitting AI as fun and rewarding as possible. Like a fun video game, along with a sacred duty to take care of users and protect their privacy.
I will also ship some form of this to end users eventually, so that it can run entirely offline, or at least without our services.
I'm thinking that systems trained with interactive learning might create an entirely new category of job, operating AI. Sure, trained domains will require less babysitting over time, but that just frees up the labor force to expand to more domains.
A little seems to help, especially when you're getting older and it's getting more painful to stare at text all day.
Too much though, and you might as well take a nap and then get some coffee. (But if you're an obsessive retard like me, you skip the nap and double the coffee. The battle waits for no one. To war!)
I do like to brainstorm super high, though. The best ideas seem to repeatedly emerge from the fog, while dead ends drift away.
But yeah, for getting things done, better to go easy on it.
I say use it if it helps you, but don't fool yourself. Keep it real.
base16-grayscale-dark
base16-grayscale-light
<3Because we can, and you will be happier. I promise.
I'm using Polly for the moment, and it provides word and viseme boundaries. Enough to animate face poses. I'm not taking full advantage of the visemes here: it's only affecting the saturation of the waveform color. Too subtle to notice, with the way things kind of flicker with this one.
I have other ideas I want to play with, but don't really have time, and I may just have to drop that line of thinking, because I plan to move to one of the wavenet implementations, which don't provide speech marks data.
I'll probably throw away the waveform. I'm just looking for ways to use the audio to pace the text output, and also to tie the speech and text together as much as possible, so there's never a gap between them for your focus to jump between.
( edit: oh, with dark and light themes! )
Fuck it. Slater is magic software...
and it's interface is like a tiling 'window manager' for any number of screens (anything with a browser), for threads of interaction, that are modeled as a story, like a conversation... but composes natural language with rich controls, forms, and stateful widgets. and super monochrome... but also controls your browser and TV and stuff...
I'm really inspired by the design work on the LightPhone 2, btw.
Super lame, early video: https://youtu.be/_m1zEbQieYQ
I might play with it this morning, with Choo!
Slater Systems LLC
Wish us luck (Slater and me)
> what is one mainstream or "consensus" view that you absolutely agree with? (This is our version of a “trick” question, reversing the now-fashionable contrarianism.)
I searched Google for variations on the contrarian inverse of the question, but couldn't find anything that seemed relevant.
Can you help me?