Currently, only if you take a photo yourself. After the release of a new model, it also will be able to see what is happening in real time.
@testingcatalog
-
Critical assessment of open-source AI model performance
By
–
It may have some use in case they will make browsing available. Currently, it is purely targeting privacy concerned userbase I am also still not confident in the performance of OS models in terms of quality. Their responses are often way too off.
-
DuckDuckGo launches privacy-focused AI chat with multiple LLM options
By
–
BREAKING: DuckDuckGO released its privacy-focused AI Chat solution. Available models:
– GPT-3.5 Turbo
– Claude Haiku
– LLAMA 3 70B
– Mixtral 8x7B -
Perplexity Adds Audio Mode for News Discoverability
By
–
BREAKING: @perplexity_ai now has an audio option on its Discover feed. There you can listen news stories in a podcast format and navigate between them Audio mode is also working in a background mode and in case the app is minimized.
-
Clarification on current AI model capabilities
By
–
It is the same model as what is accessible now. It cannot sing or change its voice yet
-
Using current AI voice models for pronunciation correction
By
–
Yes! You can already ask it to listen to how you pronounce words and correct you. But you don't need a new UI or a new voice model in order to do it. It is already possible with the current app
-
ChatGPT memory capacity and UI update explained
By
–
ChatGPT will now show a new UI in case its Memory is getting full. Currently, it can remember up to 2000 tokens and you will see this warning after 91% of the capacity is occupied. h/t @btibor91
-
Custom GPTs as AI-powered teaching tools
By
–
Yeah, and also for loads of teachers & coaches. Someone can just create a good custom GPT which could perform at the level of a human teacher. But it probably would require GPTs to obtain memory access
-
New use-case for ChatGPT voice UI
By
–
My favourite new use-case so far with a new ChatGPT voice UI 🤖
— 🚨 AI News | TestingCatalog (@testingcatalog) 7 juin 2024
* Still with the same 4o model underneath pic.twitter.com/SHJ28gewT2My favourite new use-case so far with a new ChatGPT voice UI * Still with the same 4o model underneath
