Investors bet big on new superintelligent AI startups
Reflection AI has emerged from stealth mode to join Safe Superintelligence (SSI) in the direct pursuit of superintelligent AI systems, marking a shift from more commercially oriented AI development.
Reflection AI aims to create superintelligent autonomous systems, according to the company’s website. The newly revealed startup believes that solving autonomous programming represents the key to achieving general superintelligence.
Their theory suggests that once a system masters autonomous programming, advancement in all other computer-aided tasks would naturally follow. The company combines reinforcement learning with large language models, working to give LLMs the same level of autonomy that AlphaGo demonstrated in the game of Go.
The startup brings significant expertise to this challenge. Over the past decade, team members have helped develop some of AI’s most advanced systems, including AlphaGo, AlphaZero, and GPT-4.
Ad
Second player enters superintelligence race
Reflection AI joins Safe Superintelligence (SSI), founded by former OpenAI chief researcher Ilya Sutskever, in the direct pursuit of superintelligence. SSI is reportedly in talks for funding that would value the company at up to $30 billion, despite having no near-term revenue prospects.
Bloomberg reports that Reflection AI has already secured $130 million in funding at a $555 million valuation. Investors include Reid Hoffman, Scale AI CEO Alexandr Wang, SV Angel, and Nvidia’s venture capital division.
Both companies are taking an unusual approach by focusing directly on superintelligence rather than developing commercial products first. This strategy appears to align with Sutskever’s vision – he left OpenAI in May 2024, reportedly due to the company’s increasing focus on commercial interests.
Diverging paths to advanced AI
Meanwhile, OpenAI has shifted its public messaging. The company now speaks less about superintelligence and has begun distancing itself from artificial general intelligence (AGI). While AGI remains OpenAI’s stated goal, the company now describes it as a gradual evolution rather than a specific milestone to be achieved.
This shift reflects a broader industry trend. Anthropic has moved away from the term AGI in favor of “powerful AI”. Both companies are focusing on what the industry calls “reasoning models,” the latest scaling paradigm that also relies heavily on reinforcement learning. This approach has shown particular promise in mathematics and coding, where clear right and wrong answers make reinforcement learning especially effective.
Recommendation
Business Insider reports that Thinking Machines Lab, the new venture from former OpenAI CTO Mira Murati, is in talks to raise $1 billion at a $9 billion valuation. The potential deal highlights a growing pattern in AI funding: early-stage startups, especially those founded by OpenAI alumni, are increasingly launching directly into unicorn territory.
Unlike Reflection AI and SSI, Thinking Machines Lab avoids discussing superintelligence or AGI in its announcement. Instead, the company emphasizes scientific collaboration, human-AI cooperation, adaptable AI systems, and an empirical, iterative approach to AI safety. The company plans to publish technical blog posts, papers, and code to improve public understanding of AI.
