The Daily AI Briefing - 14/04/2025

Welcome to The Daily AI Briefing, here are today's headlines! In today's episode, we're covering some major developments in the AI industry. Safe Superintelligence Inc. secures a massive $2 billion funding round, former OpenAI employees push back against the company's for-profit shift, AI demonstrates superior tuberculosis diagnosis capabilities, and several new AI tools hit the market. Let's dive into these stories and more. First up, Safe Superintelligence Inc., or SSI, has just raised an astounding $2 billion at a $32 billion valuation. Co-founded by former OpenAI chief scientist Ilya Sutskever, SSI has quickly become one of the highest-valued startups just months after its launch. The funding round was led by Greenoaks contributing $500 million, with Lightspeed Venture Partners and Andreessen Horowitz also participating. Reports indicate that both Alphabet and Nvidia are backing SSI as well, though their investment amounts remain undisclosed. What's particularly interesting is that SSI has achieved this valuation sixfold increase since September without a concrete product plan. The company is focused on building "superintelligence" that goes beyond human-level AGI while ensuring "safety always remains ahead." Sutskever has told investors that the company has "identified a different mountain to climb," suggesting a unique approach to AI development that has clearly captivated major investors. Moving on to some internal conflict at OpenAI, twelve former employees have filed a proposed amicus brief supporting Elon Musk's lawsuit against the company. These former staff members, who held technical and leadership positions between 2018 and 2024, are challenging OpenAI's shift away from its nonprofit origins. Their brief argues that if OpenAI's nonprofit wing gives up its controlling stake in the business, it would "fundamentally violate its mission statement" and "breach the trust of employees, donors, and other stakeholders." One former employee, Todor Markov, who now works at Anthropic, described CEO Sam Altman as "a person of low integrity" who used the charter merely as a "smoke screen" to attract talent. The group collectively argues that maintaining the nonprofit structure is essential to ensure AGI benefits humanity rather than serving narrow financial interests. In healthcare news, AI is making significant strides in tuberculosis diagnosis. A new study led by Swiss researchers from Lausanne University Hospital has demonstrated that AI can diagnose pulmonary tuberculosis more accurately than human experts. The system, called ULTR-AI, was trained to read lung ultrasound images from smartphone-connected devices and uses a combination of three different models to merge image interpretation and pattern detection. When tested on 504 patients, 38% of whom had confirmed TB, the AI achieved 93% sensitivity and 81% specificity, outperforming human experts by 9%. The system can identify subtle patterns often missed by humans, including small pleural lesions invisible to the naked eye. With tuberculosis cases on the rise and diagnostics often scarce in low-resource settings, this AI system could revolutionize TB testing by providing faster, cheaper, and more scalable diagnostics. Several new AI tools have been released recently. ByteDance has launched Seed-Thinking-v1.5, a reasoning AI that reportedly outperforms Deepseek R1. Writer has introduced AI HQ, an end-to-end platform for building and supervising AI agents. Amazon has released Nova Sonic, a speech-to-speech AI on their Bedrock platform. And Google has integrated Gemini into Sheets, providing access to its AI models directly within spreadsheets. In other news, Meta's unmodified release version of Llama 4 Maverick has appeared on LMArena, where it ranked below several months-old models, including Gemini 1.5 Pro and Claude 3.5 Sonnet. DeepMind CEO Demis Hassabis has shared plans to combine Gemini and Veo models into a unified omni model with better world understanding. N