In these increasingly complex and complicated times of science, technology, artificial intelligence, those across our industries find it incumbent and expensive. Matt Rogers, co-founder of Nest, is intimately familiar with these challenges. He warns that despite their success, even the largest companies need to learn how to truck in a world of total chaos. Evaluating and testing AI models, especially those built by OpenAI and Anthropic, could already be cost-prohibitive. For instance, the cost to evaluate OpenAI’s o1 reasoning model stands at $2,767, while benchmarking Anthropic’s Claude 3.7 Sonnet “hybrid” reasoning model on the same tests costs $1,485.35.
Rogers, who has witnessed the transformation of the tech landscape, noted the current dynamics: “Nest is not necessarily doing everything that I set them out to do years ago.” His outlook is emblematic of the changing culture in tech companies as they attempt to compete and make bottom lines.
The fiscal impact of AI goes deeper than model testing. OpenAI’s o1-mini, a smaller version of its models, has an evaluation cost of $141.22. By contrast, Claude 3.7 Sonnet, its non-reasoning predecessor, only costs $81.41 to use for an evaluation. These numbers underscore the massive investment that nonprofits have to make to remain leaders within the fundraising ecosystem and cutting-edge AI technology.
Beyond the dangers created by blind AI tests, harmful tech is enabled by terrible executive decisions. Just last week, former President Donald Trump signed an executive order supporting coal as a power source for data centers. Upholding this decision is a clear bias towards fossil fuels. It couldn’t come at a better time. Companies are already starting to shift their priorities to sustainability and renewable energy.
Against that backdrop, Sarah Wynn-Williams, Facebook’s former head of global public policy, recently delivered testimony before the U.S. Senate. Her experiences and comments highlight the steep regulatory ride ahead for big technology companies as they deal with both public pushback and government intervention. Wynn-Williams went on to write a memoir detailing her time at Facebook. Her insights provide continuing inspiration for the movement to change corporate governance in the tech sphere.
Karyne Levy, who was deputy managing editor at Protocol, became deputy managing editor at TechCrunch. Her job gives her a front row seat to the most important stories of the industry’s biggest advances. She is an excellent resource to help audiences navigate the evolving landscape and latest trends.
Mira Murati’s new AI venture, Thinking Machine Labs, has made strides by hiring prominent advisers Bob McGrew and Alec Radford. Their expertise could go a long way in ensuring that AI development in the company is responsive to its innovative ambitions and respectful of peoples’ rights and needs.
Under normal circumstances, this week’s emphasis on technology’s intersection with luxury would have everyone buzzing. NetJets, the private-jet company owned by Berkshire-Hathaway, just accidentally published their special guidelines for flight attendants on how to serve their chief executive passenger Elon Musk. To be clear, Musk favors very efficient travel. He’s got an operation to run, and in the simplest terms, he just wants to fly as fast and as direct as possible.
Together, these developments reflect a larger movement. Companies are going all-in on these new, disruptive technologies even as they wrestle with the cost of operations. The costly underlying infrastructure that powers AI systems further compounds the fiscal strain on companies, including startups with a mission to innovate.
Leave a Reply