In the ever-evolving world of AI, a fascinating phenomenon is unfolding: the rise of AI labs with diverse ambitions and unclear financial goals. This trend is particularly intriguing when considering the development of foundation models, a unique moment in the history of AI companies.
The scene is set by a generation of industry veterans, once key players at major tech companies, who are now striking out on their own. Accompanying them are legendary researchers, brimming with experience but shrouded in ambiguity when it comes to their commercial aspirations. The potential for these new labs to become giants like OpenAI is undeniable, yet an equally intriguing possibility is that they might choose to focus solely on research, unburdened by the pressures of commercialization.
This ambiguity has led to a fascinating question: how do we gauge the commercial intent of these AI labs? To address this, I propose a sliding scale, a five-level system that measures ambition rather than success.
Level 5: These are the AI giants, the OpenAIs and Anthropics of the world, raking in millions daily.
Level 4: Here, we find those with a detailed, multi-stage plan to dominate the AI market and become the richest on Earth.
Level 3: This level is occupied by labs with many promising product ideas, though these remain shrouded in secrecy, to be revealed in due time.
Level 2: A level of uncertainty, where only the outlines of a concept or plan exist.
Level 1: A unique perspective, where true wealth is found in self-love and not in monetary gains.
The big names, as expected, occupy Level 5. But the new generation of labs, with their big dreams and unclear ambitions, add an intriguing layer of complexity to this scale.
The beauty of this scale is its flexibility. The people involved in these labs have the freedom to choose their level, and with the current influx of AI investment, they are unlikely to face scrutiny for their business plans. Even if a lab is primarily focused on research, investors are happy to be a part of the journey. After all, not everyone is driven by the desire to become a billionaire, and Level 2 might offer a happier life for some.
Now, let's delve into some of the biggest contemporary AI labs and see where they stand on this scale.
Humans&: This week's AI news sensation, Humans& has an intriguing pitch for the next generation of AI models, focusing on communication and coordination tools. However, despite the glowing press, the team has been vague about how this translates into monetizable products. They've hinted at building an AI workplace tool, replacing products like Slack and Google Docs, but the specifics remain elusive. This level of detail, though, is enough to place them at Level 3.
Thinking Machines Lab (TML): A challenging one to rate! With a former CTO and project lead for ChatGPT at the helm, and a $2 billion seed round, one would assume a clear roadmap. But recent events, including the departure of CTO and co-founder Barret Zoph, along with several other key employees, suggest a shift in direction. It's as if they realized their Level 4 ambitions might be better suited to Level 2 or 3. While there's not enough evidence for a downgrade yet, it's a close call.
World Labs: Led by Fei-Fei Li, one of the most respected names in AI research, World Labs has an impressive track record. In 2024, they raised $230 million for a spatial AI company, which might have placed them at Level 2 or lower. But a lot can change in the fast-paced world of AI. Since then, they've released a full world-generating model and a commercialized product, addressing the demands of the video game and special effects industries. This progress suggests they might be a Level 4 company, soon to graduate to Level 5.
Safe Superintelligence (SSI): Founded by former OpenAI chief scientist Ilya Sutskever, SSI seems like a classic Level 1 startup. Sutskever has gone to great lengths to keep SSI away from commercial pressures, even turning down an acquisition offer from Meta. With no product cycles and a focus on scientific research, he's raised an impressive $3 billion. However, the AI world is unpredictable, and it would be unwise to completely rule out SSI's potential commercial impact. Sutskever's recent comments suggest SSI might pivot if research timelines change or if the most powerful AI should impact the world.
This sliding scale offers a unique perspective on the commercial ambitions of AI labs. It highlights the diversity of goals and strategies in the AI industry and the potential for these labs to shape the future of AI in unexpected ways.
So, what do you think? Are these labs truly trying to make money, or are their ambitions more nuanced? The floor is open for discussion!
