Tis the season to be jolly as Google and ChatGPT-maker OpenAI release a bounty of exciting new AI features to surprise and delight users throughout the holidays.
Ask AI Santa anything
ChatGPT’s Advanced Voice feature just got a whole lot merrier with the introduction of “Santa Mode” which enables users to chat with the big man himself. Now you can ask Kris Kringle about his naughty or nice list, talk to him about what you really want out of life, and go to sleep with tales of sugar plum fairies dancing in your head.
To activate, tap the microphone icon on the bottom right of the mobile app, then tap the filter icon in the upper right to choose a voice. Toggle to “Santa: Merry and Bright” and ask for a story. Sleigh bells ring as Santa warmly greets with a Ho Ho Ho.
Play AI guessing games
If you’re a ChatGPT Plus subscriber, you can now turn on “Vision” and “Screen Share” in the mobile app to let ChatGPT see you, your surroundings and your phone screen.
OpenAI demoed this feature as part of its “12 Days of OpenAI” in which its Chief Product Officer Kevin Weil introduced ChatGPT to his team members, then asked it to recall who was wearing reindeer ears and who was wearing a Santa hat. A fun guessing game to try at home.
But beyond mere play, ChatGPT’s new vision capabilities enables it to become a more effective coach by talking users through steps in cooking up a holiday feast and choreographing those after dinner dance moves. And with screen share, users can show it something in another app on their phone and ask what it means and how to respond, like a text from someone who is running late for the party.
Make a holiday AI movie
OpenAI has also made its AI text-to-video generator Sora available to ChatGPT Plus subscribers who can now log into Sora.com to create up to 50 five to 10-second 480p to 720p videos each month.
A simple prompt like “an orange cat pastry chef baking a cake in a holiday themed kitchen” can produce an adorable clip like “Festive Feline Baker” which is currently being featured in the community thread on the Sora site.
In its current form Sora is similar to other AI video generators like Midjourney, Runway and Synthesia, but once integrated into the ChatGPT app, the chatbot is expected to become truly multimodal.
Chat with AI podcast hosts
Among the many new features Google announced this week with the introduction of Gemini 2.0, its next-generation model for the creation of multimodal AI agents, is the hotly-awaited update to NotebookLM, a popular tool which enables users to upload up to 50 sources including PDFs of manuals and links to websites and YouTube videos, to generate an audio overview where two AI podcast hosts discuss insights from the materials. Now, users can select a “Join” button to ask questions of the podcast hosts in the midst of the show. Available to trusted testers and rolling out to Google One users.