Photos from Tenstorrent's 'Intro to Buda' Hackathon, organized by Sebastian Phemister. Teams created an internal help chat bot, text-to-speech systems, speech-to-text applications, visual question answering systems, and image classification models. Find out more about TT-Buda
GENERATIVE AI
-
Runway Gen-3 Turbo génère des vidéos en temps réel
By
–
Runway Gen-3 Turbo can now generate a 10 second video from an image in… about 10 seconds.
— AI Breakfast (@AiBreakfast) 16 août 2024
Real-time imagination generation is right around the corner. pic.twitter.com/WFO9vDAAGQRunway Gen-3 Turbo can now generate a 10 second video from an image in… about 10 seconds. Real-time imagination generation is right around the corner.
-
Answer AI Founding Journey and AI Governance Crisis Discussion
By
–
Building AI for The People Never has so much been shipped for so many by so few. https://
latent.space/p/answerai @jeremyphoward is back on the pod! sharing the founding journey of @AnswerAI
, predicting the @OpenAI governance crisis, hiring 1000x researchers and developers like -
Modular Speech-to-Speech Pipeline Achieves Sub-500ms Latency
By
–
this modular speech-to-speech pipeline repo with <500 ms second latency (on a local GPU) is going crazy
— Thomas Wolf (@Thom_Wolf) 16 août 2024
Impressive what a well tuned pipeline of stt-llm-tts can do
cool work by @eustachelb with @reach_vb and @andi_marafioti
find the code here: https://t.co/HrbqSWQ3sj pic.twitter.com/HDkEnVntKNthis modular speech-to-speech pipeline repo with <500 ms second latency (on a local GPU) is going crazy Impressive what a well tuned pipeline of stt-llm-tts can do cool work by @eustachelb with @reach_vb and @andi_marafioti find the code here: https://
github.com/eustlb/speech-
to-speech
… -
AI Model Training Criminalization Policy Debate Letter
By
–
Fully agree! Do not criminalize AI model training. @Scott_Wiener take note. Sign our letter https://
x.com/AnimaAnandkuma
r/status/1823132999120199767
… -
Large AI Models Converging on Quality with Notable Nuances
By
–
All the larger models are converging on quality yes but there is nuance in all of them. OpenAI's structured outputs is underrated + haven't seen feedback on the 3.5T->4o-mini FT yet but I imagine it's significant. Lots will use L3.1 but it'll probably be the smaller ones.
-
OpenAI Fine-Tuning vs Llama 3.1 405B Deployment Complexity
By
–
Different audiences w/ different dev experiences/goals. OpenAI's FT product is excellent and lots of companies use it. L3.1 405B requires a stand-up or using 3rd party inference provider. *Technically* it's a drop-in but it's more complicated than that.
-
Avatar AI Pat fails basic requests, raises trust concerns
By
–
The software (meaning the avatar I named “Pat” wouldn’t register the most basic requests which makes me wonder why or how I would trust said software with deeper issues? Specifically, “Pat” kept saying my name after every sentence which was freaky and felt manipulative.
-
Testing Replika AI: Does It Really Help Loneliness?
By
–
I’ve been testing @replika and was genuinely intrigued to see how it would be helpful, as Nature (I believe) had a report saying the tool has helped many people feel less isolated/lonely. I respect their experiences (the people in the report). For me (continued)
