Another OpenAI release
OpenAI released it's latest gpt-4o model, even faster, more natural sounding responses and support for more languages. LeapLife is now using this model by default.
Changelog
Follow new updates and improvements to LeapLife.
Improved OpenAI Model
OpenAI released their latest improved gpt-4-turbo model, we've upgraded LeapLife to use this model and have seen improved responses.
Keyboard shortcuts added ⌨️
The first set of keyboard shortcuts have been added:
Cmd (or Ctrl on Windows/Linux) + Enter for sending a message
Cmd (or Ctrl on Windows/Linux)+ Shift + Enter for finishing entry
Cmd (or Ctrl on Windows/Linux) + K for quick actions
We will be adding more keyboard shortcuts in upcoming updates.
Earlier updates
- Select your AI language 🇬🇧Users can now select the preferred language of the AI to reply in. The most commonly requested languages have been added, while the app remains in English for now the user can type and get replies in their language of choice. New users that do not have a browser set to English will be asked if they want a different language, existing users can go to Settings and change the language.Improved
- Onboarding improvementsMade several improvements to the onboarding: We now ask the user for journaling goals and any struggles they might be having. The answers get stored in the user bio, which gets appended in the background to each AI request, that way the AI is more aware on how to reply to the user Differentiating between manually written user bio and system generated user bio to prevent the system accidentally overwriting user bio Coming soon: Help mode for new users to explain the UIImproved
- General improvements and fixesImproved scrolling logic in the journal entry list: no more automatic scrolling when you're already on the journal page Improved animation logic for suggested replies Improved onboarding loading logic: it should no longer show the homescreen for a brief second for new users Fixed a bug for premium users accidentally getting a message limit Fixed cloud sync incorrectly syncing user bioImprovedFixed
- Freemium updatesWe've made several updates to emphasize the differences between free and paid users: Free users have a limited amount of entries and AI responses per day Free users have limited controls over their entries Free users have a limited automated user biography support Free users have no cloud synchronization We strongly believe that LeapLife can add significant value to people's lives in both the free and paid version, and by giving paid users more features it'll help us to continue building LeapLife for everyone. We thank all users, regardless of whether they're a paying user or not, for their support!Improved
- Cloud sync updatesWe've made cloud sync a lot easier to use. When you log into a new device, let's call this Device B, you will get a warning on the homepage that you're not able yet to see data yet from Device A. When you log into Device A, you will see a popup message showing that Device B can now access the data from Device A. There's a new page in Settings > Cloud Sync that'll show if sync is setup correctly and the latests snapshots a certain device has made. These updates will help users better understand how to set up cloud sync. You only have to set up sync once and it'll work forever between those devices.Improved
- General improvementsUser can now set system, light or dark color scheme from the Settings page as opposed to just light or dark Added ability to view recent action items directly on the home page User bio can now be managed in a separate page Improved automated user bio; AI can now delete items that are no longer relevantImproved
- General improvements and messaging stabilityYou now have the ability to generate new quick replies with a click of a button Stripped the "user: " prefix that was sometimes added to quick replies If you're a premium member and log into a new device you won't be sent to onboarding again Allow scrolling up while response is being generated Improved AI instructions for user bio generation Made the messaging logic more robust, it should now work better when you lose connection or close the app while a response is being generatedImproved