words get in the way (9.29.23)

Remember back in the day when the term ‘AI’ didn’t dominate every professional headline or industry earnings call? The good old days when AI wasn’t a clickbait bingo buzzword, but simply a complex, evolving technology being tested and utilized by different industries as a type of ‘spine’ within tools and technology to understand what the ability to learn and evolve with adaptability? Who would have thought in some ways 2020 and 2021 were simpler times.

?

Fast forward to today, when AI-powered tools and technologies seem to make leaps and bounds on a weekly basis. This week, OpenAi announced new ChatGPT upgrades coming soon to both Enterprise and ChatGPT Plus users that makes the movie ‘Her ’ feel more and more like a feasible way of living thanks to new multimodality offerings. ChatGPT’s new ability to process images and voice builds on its ability to process text and give users new, more intimate ways to engage with ChatGPT. How does it work?

?

VOICE : Users can enable and choose between 5 different voices and use the feature to turn ChatGPT into an assistant by engaging in back-and-forth questions and conversations with the tool. Think of engaging with it like how you currently do voice assistants like Siri, Alexa, etc.

IMAGES : Users can utilize images on their phones or images they’ve taken (with and without text in said image) to ask ChatGPT questions about the image, including specific parts of the image (think: “is that bright spot in the top left corner I’ve circled in the night sky a star, an airplane, or a spaceship?”). Think of it as a potentially more useful version of Google Lens , which lets you search what you see.

?

Both capabilities will also be made available for developers through APIs in the future, which should lead to some interesting applications. ChatGPT also brought back its browsing feature this week for Plus and Enterprise users to enable users to find-real time information via the internet, with answers coming with direct links to sources. Internet access, and its ability to deliver real time, real responses, gives ChatGPT even more of an edge for personal assistance and potential user utility.

?

If ChatGPT can now somewhat ‘see’ and ‘talk’, what does that mean for how consumers will engage with AI?? Could synthetic voices and image understanding unlock opportunities to help humans improve and enhance how they communicate with others, or provide outlets for sharing their innermost thoughts (think: potential use cases for therapy ?)? Or could it be detrimental, replacing human connection and providing false comfort (and potentially false information or impersonation)? As we see guardrails continue to be attempted and rigor put behind these tests, it will be fascinating to see what use cases are uncovered, and what actual consumer adaption of these tools become. For example, for those with smartphones, smartwatches, and smart home devices, think about what you THOUGHT you would be using voice assistants for before investing in them, and the reality of how, when, where, and why you use them.

?

It's not if you build it, consumer usage will come. It’s more if you build it, and if you show and deliver true consumer utility, consumers will show interest in testing and potentially adapting it.

?

And it’s not just ChatGPT trying to gage opportunity and interest. This week Meta announced that the Meta AI assistant , an “advanced conversation assistant,” is coming to WhatsApp, Messenger, and Instagram, along with dozens of AI characters based on celebrities like MrBeast and Tom Brady. Leveraging a partnership with Microsoft’s Bing to provide real-time web results, Meta’s goal is for the assistant to provide real-time information personified. Long term, Meta is planning on making AIs for businesses and creators available , as well as releasing an AI studio for people and developers to build their own AIs. Of course, even with the touted security guardrails, consumers are likely to have concerns. But will the curiosity of what it’s like to engage in a longer-form, intimate conversation with a celebrity trump questions around data privacy? We will see.

?

With CES only a few months away, it will be interesting to see how far chatbots like ChatGPT evolve by then, and how we see them show up (I’m going to guess we’re going to see ChatGPT and its competitors being built into hardware at CES, comparable to how in 2020 we saw countless announcements about voice assistants like Amazon’s Alexa and Google’s Google Assistant being built into smart speakers, automobiles, smart home technology, wearables, etc.).

?

While most don’t have access to this tool, would love to hear from you: How (and if) do you see voice and image-driven AI chatbots and tools enhancing your everyday life?