The current landscape for AI companies developing their own foundation models is unprecedented, marked by a dual influx of industry veterans venturing solo and legendary researchers with often ambiguous commercial aspirations. This unique environment creates a scenario where some nascent labs could evolve into OpenAI-sized powerhouses, while others might prioritize pure research without significant commercial pressure. Consequently, it has become increasingly challenging to discern which of these new entities are genuinely focused on generating revenue.

To simplify this complex situation, a five-level sliding scale has been proposed to measure the commercial ambition of AI companies, irrespective of their current financial success. This scale aims to quantify ambition, not achievement. Level 5 denotes companies already generating millions daily; Level 4 signifies those with a detailed multi-stage plan for immense wealth; Level 3 applies to companies with many promising product ideas yet to be fully revealed; Level 2 represents those with only the broad outlines of a conceptual plan; and Level 1 embodies a philosophical approach where 'true wealth is when you love yourself.'

Established AI giants like OpenAI, Anthropic, and Gemini clearly sit at Level 5. However, the scale becomes particularly insightful when applied to the new generation of labs emerging today, whose grand visions often come with less transparent commercial objectives. The substantial capital flowing into AI currently allows these founders and researchers to choose their desired level of commercial engagement, as investors are often content simply to be involved, even if the lab is primarily a research project. This flexibility means that individuals less motivated by billionaire status might find greater contentment at Level 2 than at Level 5.

However, the lack of clarity regarding an AI lab's position on this scale is a significant source of drama within the industry. A prime example is the anxiety surrounding OpenAI's transition from a non-profit, which spent years operating at Level 1, to a Level 5 commercial entity almost overnight. Conversely, Meta's early AI research might be seen as operating at Level 2 when the company's true ambition was closer to Level 4.

Among the contemporary AI labs, "Humans&" recently made significant news and partly inspired the creation of this ambition scale. Its founders present a compelling vision for next-generation AI models, emphasizing communication and coordination tools over traditional scaling laws. Despite glowing press, Humans& has remained somewhat guarded about how these innovations will translate into monetizable products. While they aim to build AI workplace tools that replace and redefine existing solutions like Slack, Jira, and Google Docs, the specifics of a "post-software workplace" remain somewhat unclear. Given their commitment to building specific, albeit vaguely defined, products, Humans& can be placed at Level 3.

"Thinking Machines Lab" (TML) presents a more complex case for rating. With a former CTO and project lead for ChatGPT raising a $2 billion seed round, one might initially assume a clear Level 4 roadmap. However, recent events, including the departure of CTO and co-founder Barret Zoph and at least five other employees citing concerns about company direction, have introduced uncertainty. Nearly half of TML's founding executives have left within a year, suggesting that their initial Level 4 plan might not have been as solid as perceived, potentially situating them closer to Level 2 or 3. Despite these challenges, there isn't yet sufficient evidence to warrant a definitive downgrade.

"World Labs," founded by Fei-Fei Li, one of the most respected names in AI research known for the ImageNet challenge, initially seemed like it might operate at Level 2 or lower, given her academic background and the focus on spatial AI. However, over the past year since raising $230 million, World Labs has rapidly progressed, shipping both a full world-generating model and a commercialized product built upon it. With clear demand emerging from the video game and special effects industries, and no major labs offering competitive solutions, World Labs appears to be a strong Level 4 company, with potential to soon reach Level 5.

"Safe Superintelligence" (SSI), established by former OpenAI chief scientist Ilya Sutskever, exemplifies a classic Level 1 startup. Sutskever has intentionally insulated SSI from commercial pressures, reportedly even rejecting an acquisition attempt from Meta. The company avoids product cycles and, apart from its foundational superintelligent model currently under development, seems to lack any immediate commercial product. Despite this non-commercial stance, SSI raised an astonishing $3 billion, reflecting Sutskever's primary interest in the scientific advancement of AI. This genuinely scientific project is at its heart.

However, the rapid pace of the AI world means it would be premature to entirely dismiss SSI's commercial potential. Sutskever has acknowledged that SSI might pivot under certain conditions, specifically "if timelines turned out to be long, which they might" or if "there is a lot of value in the best and most powerful AI being out there impacting the world." These statements suggest that depending on research outcomes, whether highly successful or unexpectedly prolonged, SSI could swiftly move up several levels on the ambition scale.