This is obviously a lie. If this was true, all the inference provider companies would go to zero. I support open-source as much as the next guy here, but it's obvious that the local version will be slower or break more often. Like, come on guys. Be real.
To illustrate this, M4 Max chips do 38 TOPS in FP8. An NVIDIA H100 does 4,000 TOPS.
Prakash if you're going to bot our replies, at least make it believable.
I am not Parkash. Just check my profile. I am not a bot. github.com / razem-io. I checked his youtube videos. He seems to lack presentation skills but his app is very usable at its current state.
I have both apps open. The STT seems to be faster with VoiceInk. Like it is instant. I can send you a video if you want.
I am sorry. I did not want your product to look bad. You are right you still need to offload the llm part to openrouter and the like if you want this to be fast too. However, having the ability to switch AI on/off on demand and context aware with custom prompts is perfect.
It can use ollama too. Yes this will be much slower but local. Best off both worlds. No subscription, even if you use cloud ai.
I mean, it's just doing STT, it's not trying to run inference on a frontier model, and if you have to connect to a remote datacenter vs your own machine that could add latency that slows down the experience. There are also tradeoffs in deployment in terms of latency vs throughput that doesn't occur on your local machine that only ever is doing STT on one thing at a time.
I'm skeptical of the claim that local is literally faster, but it's not impossible in the way you suggest.
If you watch the video you'll see that Aqua does more than just STT and why we need the frontier models. I think if you want just STT of what you said verbatim, then yea any local version of Whisper is great. If you want the enhancements shown in the video, you need the frontier models. And doing that locally is slow.
Users have different preferences on the text format they input into different apps. Aqua is able to pick up on these explicit and implicit preferences across apps – but no "open XYZ app" commands, yea
Hey thetopher, I'm the guy building the statue. My name is Pablo Peniche (@pablopeniche on twitter). I'm happy to put my name on it and answer questions.
The cost of the bigger-than-life-size statue (about 1.3x) is ~$25,000 (I'm happy to show you the receipts)
Shipping is ~$6,000 (it's "art" shipment so it's more expensive).
+ Installation at the park
+ Permits from city gov
+ Costs of the unveiling event
A 501c3 (not my own, but a tech-culture magazine "The Mars Review of Books") is sponsoring the project, so yea all donations are tax deductible.
I've personally paid ~$10,000 (again, I'm happy to show receipts) to get the design of the statue + website - and I made it all open-source so anyone can 3D print their own statue, for free.
Hi Pablo, I agree w/ thetopher that this needs more transparency. If you're happy to show your receipts to date, you should do so on the website. Also, offer a way to contact you (email, phone, something) that isn't twitter.
Aloha Pablo. Aaron was a mentor to me and one of my "heroes" (as much as one can have heros). My greatest regret in life is not being brave enough to stand up with him when he was under attack. I've cried about that often.
