The Big Idea: A Cinderella Story for Sam, but potential trouble for OpenAI
The Sam-derella Altman saga, a true Silicon Valley fairytale, just got its happily-ever-after moment. OpenAI rose to tech royalty with Microsoft's billion-dollar backing and ChatGPT's viral success. On Friday, the chariot turned back into a pumpkin. On Wednesday morning everything is back to normal and Sam’s fairytale continues, with three fewer disgruntled board members.
The Atlantic published a deep dive into the "year of chaos" that precipitated this moment. The rest of the media wrung the weekend’s Succession-worthy drama for every possible click. And for a moment, things really looked dire for Open AI with 90% of the employees threatening to leave and reports of customers looking to jump ship.
But now Altman is back at the helm, and the board is going to get some heavy hitters, moving towards operating more like a standard tech company—without necessarily ditching the now infamous non-profit governance structure.
Still, there may be some collateral damage for OpenAI. As the Information reported, this week’s events forced many customers to start exploring other platforms—a trend we heard amongst our customers as well. Even with Sam at the helm, one of the buzzphrases of the day is “diversify your models” so you aren’t beholden to one platform.
So who are the real winners and losers in this Silicon Valley meets Game of Thrones saga?
- Sam Altman: Triumphantly returns, reshapes the board to his vision.
- Satya Nadella: For the brief moment that he had “hired” Altman, Microsoft’s market cap jumped by $63B. But having Altman back at OpenAI seems like what Nadella wanted all along, and Microsoft will likely get a board seat.
- Anthropic, Google, and Microsoft: Each got a long look from OpenAI customers this week, and we’ll likely see more model diversification moving forward.
- The Doomers on OpenAI's Board: They were never able to materialize any evidence against their claims of Altman withholding information. Plus it was just really embarrassing and put the company at risk.
- Ilya Sutskever: Probably lost his Thanksgiving appetite after trying to oust his boss and then backtracking all the way.
In terms of advice for what to do now, our advice is: Take a deep breath and enjoy the turkey. We’ll have more on model diversification after the tryptophan wears off.
CHART OF THE WEEK
Less than 10% of enterprises have reached full AI integration
After more than a decade of investing venture capital in AI and backing more than 50 AI-related companies, Coatue Management has some insights to share. Their latest study found that only about 60% of surveyed enterprises plan to adopt AI. Among those who have already started, less than 10% claim they’ve reached full adoption.
Coatue has identified several key considerations informing their AI investment approach based on how AI has impacted their venture ecosystem, modern tech stack, and the broader economy in 2023. Their take? AI is making strides at a speed incomparable to previous technology revolutions. And unlike other trend cycles, the hype around AI is justified by real-world utility.
But, with around 40% of companies still on the sidelines of AI integration, there's a vast landscape of untapped potential.
In order to be successful in the current AI wave, it is crucial to have a deep understanding of the various layers of technology involved, how they work together, and what sets them apart from previous technology stacks. A new generation of AI Ops tools has emerged, making it significantly easier to train, refine, and deploy models, as well as to build practical applications on top of them. With the introduction of a dynamic AI Ops layer, approximately 30 million developers around the world now have access to AI, which has the potential to drive a revolution in application development.
Coatue’s advice to startups building AI applications? Move quickly and create new behavioral shifts in order to compete with incumbents. In the long run, Coatue predicts that the development of AI will become an engineering versus research challenge as companies aim to scale AI across billions of users.
We’re at an inflection point—where early adopters can become leaders, and hesitation could mean missing out on defining the next frontier of AI innovation.
A Framework for Using Generative AI at Your Company
Given the OpenAI chaos of late, now seems like a good time to talk to a lawyer. Especially if you’re bringing in outside help to build AI at your company. We’ve got Michael Rispin, GC of Sprout Social on speed dial—you might recognize him from our expert panel on Data Safety last month. He developed this a framework for navigating this question, taking a nuanced approach to integrating AI into a business model and evaluating risk without sacrificing innovation.
The flowchart above offers a basic outline for assessing the risk of a new AI vendor. Note that several paths lead toward Rispin’s three favorite words “Check with Legal.”
A key factor often overlooked in the race to adopt AI technologies is the distinction between low-risk and high-risk AI tasks. The warning is clear: Poor data management can not only derail AI projects but also devalue your company.
Master multilingual AI with a leader in AI-driven language education
Need help developing multilingual LLMs?
If you’re in NYC on Nov 30; Join Anthony Spadafino (Founder and CEO of Lingostar.ai and ex-Uber) for a special early bird workshop before the AI x Future of Work Summit to learn the ins and outs of creating multilingual chatbots.
AI DISCOVERY ZONE
Ever wish David Attenborough was narrating your life? An X user created a Python Script so you can star in your own Planet Earth.
DEEP DIVES FROM THE ARCHIVES
- Why the Anti-Open Sourcers Have AI All Wrong
- 3 Things Generative AI Can Never Do Better Than Humans
MEME OF THE WEEK
Missed last week’s issue of Build Mode? Read it here.