🌊The Long View

For any frontier tech to “cross the chasm”, it has to pass the CFO test: that is, move from being a line item in the innovation budget to the operations budget.

For startup founders, the potential of working with an enterprise customer is alluring, and it feels tempting to jump through hoops in a bid to get pilot partnerships. But rather than one-off innovation experiments, what everyone at the table should design for is the long view: how might this startup product or technology drive impact on the business in the medium to long run?

Sometimes it’s about optimizing the bottom line, or giving the business new markets to expand to — there are many reasons why larger companies want to work with startups with great ideas. For startup founders: you want to work with your pilot customers to move from the innovation bucket to operations as quickly as possible.

Even with the AI hype in public and private markets, a study from Deloitte notes that large-cap CFOs are still relatively cautious when it comes to AI. While 82% of CFOs surveyed said that they will increase spending on AI in 2024, close to 62% of respondents have only allocated less than 1% of their budget to Gen AI. A majority of them cite that they struggle with challenges such as measuring the technology’s real business value. 70% of CFOs forecast that the technology will have a strong impact on productivity.

We are still early in this technology cycle. Enterprises have to balance the tension between exercising caution and keeping up with innovation — or risk being disrupted. There’s already talk bubbling up of giants being felled.

🔥 The Latest in AI

1

Great news for the ecosystem: Several open source large language models (LLMs) were announced this week — Stability unveiled its 12B parameter Stable LM2 model, Mistral released Mixtral 8x22B, a new large language model and its latest attempt to compete with the big boys in the AI arena, and Meta confirms that its Llama 3 open source LLM is coming in the next month.

Open source is fundamental for a thriving ecosystem, as it democratizes access to foundational models, meaning developers can build on and modify a foundational model codebase without the need for capital-intensive training from scratch. 

2

AI-generated music holds staggering promise, and competition heats up. Will the ultimate test be about distribution? Spotify launches personalized AI playlists that users can build using prompts. The music app says it’s using large language models (LLMs) to understand the user’s intent, and then uses its personalization technology — the information it has about the listener’s history and preferences — to fulfill the prompt and create a personalized AI-generated playlist for the user. On the music generation front, Udio came out of stealth with $10M in seed funding from prominent investors and celebs. (From a taste-standpoint, I thought Suno’s generations are way better at this moment in time. Note: I am an investor in neither.)

3

Will we be seeing a wave of creator revenue-share models for AI? Poe introduced a new ‘price per message’ monetization model, which allows creators to generate revenue each time a user interacts with their chatbot. The race to create an “app-store” ecosystem is on.

4

Apple vs Apple? AI hardware company from Jony Ive, Sam Altman seeks $1 billion in funding: Former Apple design lead Jony Ive (behind the legendary iMac, iPod, iPhone) and OpenAI CEO Sam Altman are seeking funding for a new company that will produce an "artificial intelligence-powered personal device," according to The Information.