- odo newsletter
- Posts
- 👀 👂 👄 ChatGPT Can Now See, Hear, and Speak
👀 👂 👄 ChatGPT Can Now See, Hear, and Speak

Welcome to the odo newsletter—your weekly digest on AI for product builders.
Be in the Know 👀
ChatGPT Provides Up-to-date Information: ChatGPT can now browse the Internet and provide current information with reference links. Previously, ChatGPT's responses were limited to data up to September 2021. ChatGPT Plus and Enterprise users can start accessing this feature with GPT-4.
Spotify Allows Podcasters to Translate into Many Languages: Spotify launched a new feature that allows podcasters to translate their podcasts into additional languages in partnership with OpenAI's voice recognition technology. The feature aims to match the original speaker’s style and help increase audience size.
Meta to Launch AI Chatbots Embodying Celebrity Personalities: Meta will launch Meta AI, personal assistants that embody personalities of celebrities such as Snoop Dogg, Dwyane Wade, and Kendall Jenner. 28 different AIs will be available in beta across WhatsApp, Messenger, and Instagram in the U.S., as well as smart glasses and Quest 3. It’s Meta's way of appealing to the younger audience, especially given the success of companies like character.ai.
Mistral AI to Launch its First AI Model: A French AI startup, co-founded by Google DeepMind and Meta alums, has launched its first model called the Mistral 7B model. While it’s not technically an open-source model, it is released under Apache 2.0 license, which imposes no restriction on use or reproduction, aside from attribution.
Writers’ Strike Ends, with Mixed Implications on AI Usage: According to the new bargaining agreement, “AI can’t write or rewrite literary material, and AI-generated material will not be considered source material.” While this seems like a win for the writers, the agreement hasn’t outright banned the studio or writers' use of AI.
Amazon to Invest $4BN in Anthropic: Amazon makes big moves into AI-powered chatbots, similar to Microsoft's Investment in OpenAI and Google's in Google Bard. Anthropic has found a strategic investor that can help provide compute power and sales channels.
CIA Launches Its Own AI-powered Tool: “The tool will allow analysts to see the source of information and ask questions about what they are viewing,” says Randy Nixon, director of the CIA’s Open-Source Enterprise division. Let's hope the tool doesn’t hallucinate too much and pin down the wrong person.
News Deep Dive 🤿
👀 👂 👄 ChatGPT Can Now See, Hear, and Speak
OpenAI has started to roll out new voice and image capabilities in ChatGPT. You can take or upload a picture and ask questions about it to ChatGPT, such as taking a picture of your fridge and asking what to make for dinner. You can also converse with ChatGPT on your mobile phone, instead of typing instructions in a chat box. These features will be available to ChatGPT Plus and Enterprise users over the next two weeks.
I (Yooni) got access to the voice feature and tried it out. I asked how to kill a fly, since there's been a fly in my apartment since yesterday and it's been bothering me.
ChatGPT empathized with my situation ("Ah, flies can be quite annoying, can't they?)
It provided a few options (Fly swatter vs. fly spray).
It asked how I have been dealing with the situation. When I responded that it's just one fly, it adjusted its recommendation ("If it's just one fly, a fly swatter is probably the quickest way to get rid of it…if you don't have a fly swatter, a rolled-up magazine or newspaper could work in a pinch.")
Even in this brief interaction, I was struck by two things: 1) I was pleasantly surprised by its ability to “carry on” a conversation, as opposed to in a chat setting where it feels like a one-way conversation. 2) It felt much easier to talk to the bot as opposed to type out my request, especially for something casual like how to get rid of a fly.
We're excited to see how the image function performs!
Product Resource ☀️
Rocks, Pebbles, and Sand is a popular framework for categorizing different types of work your team does. This guide shows you common pitfalls that happen with each category of work and how to best apply the framework.
Job Postings 💼
Sr. Product Manager, Alpha (AI) at Public
Senior Product Manager, Google Cloud Generative AI at Google
Staff Product Management, AI at Walmart
VP of Product at Copy.ai
Disclaimer: The job postings listed in this newsletter are for informational purposes only. We are not endorsing these positions.
Before you go 💨
A boy was sick over 3 years and had to see over 17 doctors. Frustrated, his mom turned to AI for help, and ChatGPT suggested a diagnosis that turned out to be accurate. We’ll leave it up to you if you want to try it too!
Reply