Ron Shina
203 posts


@thsottiaux - access to Pro in the codex app
- keyboard shortcuts to navigate between prompts and responses
- metadata about a response - which model on which reasoning effort and when
- exporting a thread (prompts and responses)
English

@NotebookLM (As long as there no easy way to select a subset of notes) the ability to "open a folder" which adds the sources in its children
English

@NotebookLM - "Which Notebook has this source?" - search starting from any folder
- copy / move Sources to Notebook (this enables using Notebooks for subsets of Sources, until that feature is added)
English

@thsottiaux (Codex app) "Check for updates" or a notification one exists
English

@thsottiaux - Scrolling with the keyboard in the app
- Quickly navigating responses and prompts
- Exporting the input and output of a thread / all threads in a repo ...
English

@joshwoodward Deselecting Deep Think starts a new chat, so it would be nice to have an "Export Chat", or "Branch in new chat" (like in chatgpt) to have the content of the Deep Think chat in the sans Deep Think chat
English

Gemini 3 Deep Think can help make things. 🧠
Here's our side project: We sketched a laptop stand and Deep Think coded that into an interactive prototyping tool. We used that tool to generate a STL file, which we sent to @fleet_ai. And now I have a new laptop stand!
What will you build? See the final photos below!
Google Gemini@GeminiApp
Today, we’re releasing a significant upgrade to our specialized reasoning mode, Gemini 3 Deep Think. Deep Think is built to drive practical applications, enabling researchers to interpret complex data and engineers to model physical systems through code. With the updated Deep Think, you can turn a sketch into a 3D-printable reality. Deep Think analyzes the drawing, builds the complex shape, and generates a file so you can create the physical object with 3D printing. This is rolling out now to Google AI Ultra subscribers. Select the "Deep Think" option in the tools menu to get started. Learn more here: goo.gle/3MoiifF
English

@athyuttamre Maintaining the requested accent in English and Spanish
English

@athyuttamre running a thinking query in the background, continuing the conversation and it notifying you 🔔 there's a response and you can talk about it
English

@joshwoodward @DataChaz You can even leapfrog openai by making the list of projects / folders searchable (like Bookmarks in x) :)
English

@athyuttamre @simonw Maintaining the requested accent in Spanish and English throughout a conversation
English

@simonw We’re working on it! What are the top things you’d like to see fixed/changed?
English

@joshwoodward @DataChaz a "skip pinned chats" would also be nice
Currently, you need to remember to pin a conversation you get back to often, if you already have many pinned, to avoid having to scroll (which can incur additional loads)
English

@joshwoodward @GeminiApp (Line Breaks) one of the reasons I use ChatGPT for many tasks is how often output from gemini doesn't contain the same line breaks as text in the prompt (when asked to manipulate it in some way)
English

✅ Papercut fixed: You can now tap "Answer now" to stop @GeminiApp from thinking and answer right away (using the 3 Flash model)

English

@joshwoodward (pinned conversations) Easily changing their order - I think pinned conversations should be thought of like a playlist
In that regard, YouTube web has both "Move to top" and drag and drop for reordering faster
Google Keep also enables reordering of pinned notes
English

@dlevine815 (pinned chats)
changing their order similarly to how it's done in music playlists (no need to send a dummy message to move a pinned chat up or unpin and pin again, etc)
English

@dlevine815 - Reducing repetitions in Pulse between days ("I know, you've told me that on Wednesday.")
English

@dlevine815 - Metadata - which model configuration answered and when
When - for multiple reasons, maybe something has changed ("Did I ask before that happened?")
Model configuration -
"Was this answer given by their best model?" / Models improve over time, "let's ask again"
English

@dlevine815 - Deleting messages from the middle of a chat
Sometimes you ask for more details and the older messages become clutter
English

@joshwoodward Searching (at least) a chat using a gemini query or some better exact search for long chats with repetitions (for example, "search in selection")
English

@dlevine815 (Projects) Loading times for "... See more" on the web app and also after scrolling down the list
this leads to sending a dummy message in a chat just to have the project bumped to the top of the list (to not have to wait so long next time)
it's faster on the mobile app
English










