Olivier
@StonkyOli
Followers
724
Following
713
Media
91
Statuses
720
Terminally online. Builder of things. dm me your secrets.
San Francisco
Joined October 2021
Don't wanna be here?
Send us removal request.
Explore trending content on Musk Viewer
Musk
• 1038474 Tweets
無料クーポン
• 325220 Tweets
増量作戦
• 212459 Tweets
#ファミマ増量クランチ
• 62466 Tweets
百鬼夜行
• 53793 Tweets
智弁和歌山
• 53105 Tweets
アジアツアー
• 45070 Tweets
台風8号
• 30411 Tweets
サブスク解禁
• 19914 Tweets
#映画ふれる
• 19785 Tweets
LagunyaZAGAT PasBANGETTT
• 19591 Tweets
GALAUmu GALAUku
• 18852 Tweets
KDRT
• 18078 Tweets
#ふれる残暑見舞い
• 16130 Tweets
シャトレーゼ
• 11873 Tweets
新人さん
• 11659 Tweets
WE LOVE JOLLIBELLE
• 11371 Tweets
#INZMARATHON
• 10012 Tweets
@svpino
Devin - flat out scam / staged demo
Humane - bad software execution, over promised and underdelivered, will likely get better though
Rabbit - I think it will be a cool toy, but will likely get left behind pretty quickly, their “action model” is a facade for hardcoded selenium
7
6
319
13
8
49
@polynoamial
@hughbzhang
@summeryue0
@scale_AI
This is great but the color coding based off slight shades of a color is a bad representation
0
0
45
@daniel_nguyenx
The term LAM was coined by Silvio Savarese () in '23 where he describes them as LLMs that are able capable of performing tasks autonomously rather than just responding with text, under this definition function calling is indeed LAM
The term then became
2
1
34
Excited to share a small experiment in AI wearables: a device that transcribes conversations and identifies individual speakers, adding a much needed contextual dimension to how we interact with LLMs. Watch the video to see it in action.
6
2
31
2
1
21
100% agreed, response latency is only 1 dimension of getting a realistic conversationalist.
Knowing to stop when interrupted (and maintaining context of when this interruption happened)
Knowing when someone is "thinking" in between words, and not jumping the gun to a response
everyone's optimizing response latency, but that's rarely what breaks user experience. it's interruptibility
most can't be interrupted, and fixing text transcript after an interruption requires TTS to return word-level timestamps
TTS APIs: plz return timestamps
10
1
84
2
3
15
Me in the simulation Me at the IRL party
launching the party simulation now
119 avatars are joining as we speak
let's see how it goes 🎉
11
4
43
1
3
15
@SullyOmarr
This is indeed brutal
The most mind boggling thing is the voice-to-voice latency, teams like
@Vapi_AI
have shown you create sub sub-second pipelines
TBF there is also action calling involved on the humane pipeline, but 10 seconds just doesn't quite add up, even under poor
5
0
12
@yoheinakajima
While i love the concepts langchain has thought me; I consistently get better results "raw-dogging" the openAI API as well.
2
0
10
GPT can interpret data requests, write valid SQL queries and decide the best way to represent the data to an end user 👀
Little side project i've been working on
3
1
10
Proof that
@perplexity_ai
uses google search in the background 🙄
Websites recently de-indexed by Google are also “unavailable” on
@perplexity_ai
These same websites are available and indexed on every other search engine such as ddg/bing/etc
I also suspect they only look at
@gaganghotra_
@agihippo
Here I tried a query about a website recently deindexed from Google—note Perplexity's response: "Unfortunately, there are no search results
provided for the website godownsize.com."
1
0
1
4
0
8
100% agreed, response latency is only 1 dimension of getting a realistic conversationalist.
Knowing to stop when interrupted (and maintaining context of when this interruption happened)
Knowing when someone is "thinking" in between words, and not jumping the gun to a response
2
3
15
1
1
8
3
0
6
@SergioRocks
Someone recently told me to *NOT* use GPT if i want to become a better programmer
This was shortly after assuming I was a bad programmer because I've never worked in big tech 🤦♂️
3
0
6
0
0
6
@NickADobos
I built a GPT 4 version of this if you want to try it out: - summarization happens on h100s running mixtral for speed, but final formatting is gpt4 powered
0
0
5
@sahir2k
Rawdogging LLMs is the best way to go about it IMO
Abstractions take away too much flexibility
1
0
5
@GroqInc
I'd tell you a joke about parallelization, but it seems to lose its timing when executed sequentially
0
0
5
@dannypostmaa
Unpopular opinion but this is why typescript is so good, you can write full stack web and mobile apps in the same language
Plus static type safety and active ecosystem out of the box
4
0
4
0
0
4
@domwhyte42
@OpenAI
I built something similar a couple weeks ago, but it connects directly with a DB instead of using spreadsheets.
GPT can interpret data requests, write valid SQL queries and decide the best way to represent the data to an end user 👀
Little side project i've been working on
3
1
10
0
0
3
Alright let's build something. I'm looking for a new apt in SF and scheduling consecutive viewings in the same day that are near eachother has been a painpoint. Going to create a quick app where i can input a list of links and it will:
1. Scrape key data about the listing such
1
0
4
@CustomerIO
I’ve been looking for you all my life.
Amazing product, can never go back to sendgrid or equivalent. This is the way email systems should work ❤️
0
1
3
@Stefcodes
I’m just annoyed I have to discern between the 2 versions when debugging or looking at resources online
0
0
3
@EthanSutin
@BartronPolygon
I've been working on something similar 👀
Excited to share a small experiment in AI wearables: a device that transcribes conversations and identifies individual speakers, adding a much needed contextual dimension to how we interact with LLMs. Watch the video to see it in action.
6
2
31
0
0
3
It's rare that you find people who are literally on the same page as you ♥️♥️
Fascinating discussion on
#AI
and Humanity's future happening right now. If you have any thoughts please come and share them.
This is a totally open discussion.
4
10
78
0
0
3
@kyleshrader_
@yoheinakajima
First individual employees
Then entire departments
And ultimately entire orgs
2
0
3
0
0
3
V2
Excited to share a small experiment in AI wearables: a device that transcribes conversations and identifies individual speakers, adding a much needed contextual dimension to how we interact with LLMs. Watch the video to see it in action.
6
2
31
1
0
3
@OpenAI
Had this awful dream where the announcements were disappointing
Relieved to wake up and realize the event hasn’t happened yet lol
1
0
3
@202accepted
I think what a lot of us missed is a startup comparing themselves with google and us just accepting it
1
0
3
@makowskid
@svpino
Same experience with unusable “lag” on both windows and osx
After using gaming mice with dedicated RF dongles the difference is hard to forgive, but it’s one of those things that you don’t notice until you’ve tried it, and then you can’t go back. Much like high refresh rates or
0
0
3
@willdepue
@sincethestudy
Yeah, have been on vercel stack for 3 years. Zero issues in terms of downtime.
0
0
3
SF wide scavenger hunt happening now by an unknown group. Really well executed, it has a full blown web app and seriously fun puzzles that make you run all around the city.
1
1
3
@RampCapitalLLC
I looked up “ring” on Costco (Instacart) and you can get a diamond ring same day delivered.
What have we done.
0
0
3
@lucy_guo
@dimalepokhin
has done things i thought were impossible on webflow - specifically around animations and motion.
1
0
3
@WesRothMoney
@DrJimFan
Maybe they have an algorithm to choose relevant frames outside of the model itself 🤔
0
0
2
@petergyang
@anothercohen
I've noticed this too, using code to break up the task into smaller chunks then programmatically stitching the responses together should yield improved results here🤔
I've been finding myself wanting to use it for larger JSON and CSV transformations as well.
0
0
3
@bitcloud
On audio, monolith architecture (vs stitching 3 models together) should in theory go a long way, time to first token is most important on speaking use cases, humans speak relatively slowly
0
0
3
@MatthewBerman
We've been stitching together STT -> LLM -> TTS models to achieve voice to voice, then plopping end-of-turn detection on top of all this to identify when the user is done speaking
An all-in-one model should in theory have tons of advantages, and has the potential to be far more
0
0
2
@JoshWComeau
100% - someone should make a chrome extension that lets you change the prefix (and set a default preference). The only reason to partially select text is when you're using a different package manager.
2
0
2