Applications over AI models…Scott Belsky's copyright concerns…The new tech cold war. Here's what you missed from Cerebral Valley New York

Applications of AI

We bring you eight more conversations from the Cerebral Valley AI Summit held in New York last week.

subscribe Newcomers YouTube Channel Stay tuned for more highlights.

Lamp Eric Griman Intercom Eoghan McCabe Each spoke about how their company is currently adopting AI tools: McCabe mentioned that search augmentation generation (RAG) was a breakthrough for Intercom, while Griman described how Ramp is using AI for features like receipt mapping, transaction coding, and expense notes.

McCabe argued that incumbents facing an influx of AI-native challengers need to jump on the AI ​​bandwagon before it's too late. “You have to be prepared to put all your money on the line and take crazy risks,” he said.

Sequoia Pat Grady He argued for greater stability and consistency in the foundational model layer, which would better support founders working on AI applications. He argued that better models don't necessarily make companies better, because the models are already good enough: “I think the current capabilities are enough to build trillions of dollars' worth of new businesses.”

of belief Sarah GuoThrive Capital Vince HankesIndex Ventures Shardul Shah The day started with a temperature check of which areas of the AI ​​stack are likely to yield the most benefits. Unlike past summits where OpenAI was always the talk of the town, the trio felt that with the proliferation of competition and open source models, the real value of this wave of AI is in applications. “The majority of founders should not be working on the foundational model,” Hankes said.

Crusoe CEO Chase Lockmiller He led a breakout discussion on the need for more computing power to fuel the AI ​​boom.

Founders and investors questioned whether data centers could be built fast enough, with most already nearing capacity, and Lochmiller emphasized that more renewable energy resources are needed to meet AI demand.

Primary Tobias Citron He questioned the assumption that scale alone is the path to continued improvement in AI, suggesting that building more middleware to run AI tools more efficiently is also a big piece of the puzzle.

Scott Belsky He shared how Adobe is responding to the wave of AI, emphasizing that they only use their own models trained on licensed data to keep customers comfortable and out of court. “Customers feel at risk. They never want to be trained on their own data,” he said. He added that brands are also very fearful of copyrighted content slipping into their own material.

Belsky said flexibility will be key for everyone in managing the transition to AI: “As platforms shift and strategies change, you need to lead with adaptability,” he said.

The looming threat of China's AI development is a major concern for General Catalyst advisers. Theresa Carlson Co-founder of Lux Capital Josh WolfThey spoke with the former founder. Zoe Weinberg.

Carlson and Wolf agreed that China is six to nine months behind the U.S. in AI, and that it's important for the U.S. to continue investing to stay competitive. But Wolf disagreed with those who call AI the “new nuclear weapon” in geopolitics: “AI is much more diffuse than localized, controllable atomic matter.”

CEO of Datadog Olivier PommelComing from the New York tech industry, he argued that the future of the AI ​​industry is brighter in the Big Apple because the industry is more diverse, making it “easier to understand how to develop the right products than in the Bay Area.” He said that most of his small and medium-sized business clients are still in the experimentation stage of AI tools and have not yet developed a clear AI strategy.

On the consumer side, Grindr CEO George Allison He proposed a world in which AI helps us build stronger personal connections with friends and loved ones, especially those who don't live nearby. AI dating coaches could help us build confidence in talking to our matches, and “AI wingmen” could pre-screen potential matches to ensure romance works out. “Our artifacts will talk to each other and then relay those conversations back to us,” he said.

Source link

Leave a Reply

Your email address will not be published. Required fields are marked *