What I’m thinking about in 2024, GPT Store launch, and the rise of AI Apps:
Also AI tutors, Bitcoin ETFs, and Nividia's newest chip
Happy 2024! Over the past few weeks, I have been reflecting on the tech and venture capital landscape over the past year and what it means for entrepreneurs in 2024. I also spent time thinking about how to utilize Substack to share what I’m thinking about this year. This brings me to why I enjoy partnering with technology founders and building industry-changing companies.
I see today’s entrepreneurs as the contemporary artists of our era. I have studied fine art and design throughout my life (fine art major in college), driven by my joy of creating something new. This process later ignited my passion for entrepreneurship and computer science. Just as a sculptor shapes a block of marble, an entrepreneur must decide which strategies to prioritize and which to discard. Like artists, software developers, and technology entrepreneurs use elements from the past to shape the future, transforming the unimaginable into the inevitable.
As an Investor, I apply a design thinking mindset to partner with entrepreneurs who are customer-obsessed. New technology companies can catalyze change and promote meaningful discourse, much like a work of art.
Throughout this year, I will use this platform to explore and discuss technology advancements, platform shifts, and the overlaps between culture and technology. I’m currently excited about the following areas:
Developer tools and infrastructure that help companies adopt AI and support the growing open-source ecosystem.
Emerging model domains that integrate vision, text, voice, and more, extending beyond just language.
Enterprise applications that enhance productivity across various teams such as marketing, sales, design, finance, product, procurement, and security.
Consumer applications that foster social connection and experiences, address the loneliness and mental health epidemic. Apps and games that promote companionship, entertainment, and creativity.
Marketplaces with AI on the other side, specifically applied to services marketplaces.
Applications that introduce innovative UX/UI and redefine business models for our AI-native future.
A few times each month, I will aim to cover a mini deep dive on one of these spaces and what I’m reading.
The GPT Store & AI App Predictions:
A few weeks ago, I shared my 2024 predictions with Geekwire. My predictions centered on two key points: (1) AI meeting users where they already are, and (2) new AI apps introducing innovative UX/UI. I believe we will see various AI-native apps emerge with winners that unlock new user behavior. Which is already starting to show up in OpenAI’s GPT store.
The GPT Store launch signifies a shift away from AI’s focus on the infrastructure layer to the app layer. With foundation models now accessible via APIs and a growing open-source model ecosystem, I anticipate a surge of AI-native apps this year that gain initial distribution on the GPT Store and later evolve into robust products and companies. The winners will stand out by offering a superior user experience.
A survey, last August, revealed that while most Americans are aware of ChatGPT, few use it regularly. In August, only 18% of US adults had used it. This suggests that while in tech it feels like everyone is using ChatGPT, there is a huge opportunity for applications built on top of existing models, with improved UX/UI that attract new AI users. Vertical-specific apps that connect to the end user's workflow will likely drive adoption.
If we look back in app history, when Instagram and Snap launched in 2010 and 2011, about 30% of U.S. adults had a smartphone. By 2017 when TikTok launched in the U.S. that figure had risen to around 70%. The chart below from Coatue’s AI research report, shows we are still early in AI technology adoption, but the rate of adoption is much faster than the internet and smartphones. I expect the application layer to start to proliferate this year, and I expect it to develop faster in this era than in past platform shifts.
The application layer is where humans will interact with AI. Apps will evolve from delivering software-as-a-service (SaaS) to Intelligence-as-a-service (IQaaS), enhancing our capabilities, modifying behaviors, and automating workflows.
Incumbents will likely continue to shine in the short term because they can take advantage of their data and distribution when they deploy AI features. However, AI-native companies can build faster, are more nimble, and can attract incredible talent. AI-native apps have the potential to unlock new behaviors, much like the combination of GPS, mobile cameras, and mobile internet and app stores did for mobile apps. Now, the combination of AI models, the GPT store, and new UX/UI is poised to provide the surface for new apps that can unlock new behaviors in this AI era.
Initially, these apps may sound far-fetched and the business models may be unclear. However, these unusual ideas often unlock new user behavior. Before Uber, it was weird to get in a stranger's car, and before Snap we didn’t communicate through selfies. Major app startups from the past decade like Uber, Instagram, and WhatsApp changed user behavior and reinvented our daily expectations. We are entering a world of apps that change our expectations around personalization, provide infinite gaming experiences, generate TV shows, have AI shopping assistants, and reason and automatically complete tasks for us.
While writing this, the top 6 trending apps on the GPT store provide productivity that was unimaginable 18 months ago. Consensus has a GPT that allows users to search 200M academic papers to find science-based answers. Grimoire is a coding assistant that can create a website from a sentence. AskYourPDF Research Assistant allows users to chat with multiple PDFs, generate content, and validate citations.
While a variety of apps have launched in the GPT Store, I see potential for new AI applications outside of this platform. The GPT Store may serve as an initial entry point for users and a distribution strategy for apps. But, GPTs are limited to a chat-based UX. I believe applications that truly unlock new behavior will be multi-modal and have reimagined UX/UI for their specific use case.
What I read last week:
Will Chatbots Teach Your Children? Online learning platforms such as Khan Academy and Duolingo have introduced AI chatbots built on GPT-4. Chatbot tutors offer personalized learning experiences, that could narrow existing gaps in education, assist teachers, and even identify previously unnoticed learning disabilities.
The SEC Approved Bitcoin ETFs: Now, consumers can purchase Bitcoin from Fidelity or Blackrock, and track this ETF alongside their other investments in the same account. This simplifies and demystifies the Bitcoin investing process, potentially accelerating regulations aimed at eliminating fraud and normalizing cryptocurrency as a significant payment infrastructure.
Microsoft Copilot has more than 5 billion chats. In addition to this milestone, Microsoft announced Copilot Pro, which includes the ability to create Copilot GPTs. A Copilot app for iOS and Android was also unveiled.
Nvidia announced new graphics cards designed for generative AI applications. These new cards are tailored to run AI models without needing to send information back to the cloud.
Introducing r1, a pocket company that moves AI from works to actions: The product aims to offer an app-free online experience that can understand user intentions and behaviors. The company has already sold over 10K units.
Evolution of Ads Conversion Optimization Models at Pinterest: The Pinterest engineering team shares their experience with incorporating Multi-Task Learning and leveraging real-time user action signals to enhance their ads recommendation system.