For AI companies the Internet is too small
For AI companies the Internet is too small
AI models that cost $1 billion to train are underway
Right off the batch: 50% of Y Combinator 2024 startups built with AI
Apply for the CO— 100
The CO— 100 is an awards program run by the U.S. Chamber of Commerce that recognizes and celebrates America’s Top Small Businesses.
This exclusive list honors businesses that excel and stand out in 10 categories: Growth Accelerators, Community Champions, Micro-Business Leaders, Champions of Adaptability, Customer Champions, Global Stars, Digital Innovators, Enduring Businesses, The Disruptors, and Culture Champions.
Ten finalists will be chosen from the CO— 100, and each will receive a $2,000 award. One business will receive $25,000 and the title of “America’s Top Small Business of the Year.”
***
For AI companies the Internet is too small
In the rapidly evolving world of artificial intelligence (AI), companies are facing a new challenge: the internet may not be extensive enough to support their growing needs. As AI systems become increasingly sophisticated, they require vast amounts of data to learn from, putting a strain on the currently available public data online. This issue is further compounded by some data owners denying AI companies access to their resources.
According to industry insiders, the demand for high-quality text data may exceed the supply within the next two years, potentially slowing down the progress of AI development. As a result, AI companies are now exploring alternative data sources and rethinking their training methods.
OpenAI, the creators of the popular ChatGPT, have reportedly considered training their next model, GPT-5, on transcriptions of public YouTube videos. Meanwhile, other companies are experimenting with using AI-generated, or synthetic, data for training purposes. This approach, while still….
Originally by Deepa Seetharaman of The Wall Street Journal
***
AI models that cost $1 billion to train are underway
The cost of training advanced AI models is set to skyrocket, with Anthropic CEO Dario Amodei revealing in the In Good Company podcast that some models currently in development can already cost up to a staggering $1 billion. While current models, such as ChatGPT-4, require around $100 million to train, Amodei predicts that this figure could increase to anywhere between $10 to $100 billion within the next three years.
"Right now, 100 million. There are models in training today that are more like a billion," Amodei stated. He went on to say, "I think if we go to ten or a hundred billion, and I think that will happen in 2025, 2026, maybe 2027, and the algorithmic improvements continue a pace, and the chip improvements continue a pace, then I think there is in my mind a good chance that by that time we'll be able to get models that are better than most humans at most things."
During the discussion on the development of AI from generative artificial intelligence (like ChatGPT) to artificial general intelligence (AGI), Amodei emphasized that the progression would not occur at a specific point. Instead, he described it as a gradual process, with models building upon the advancements of their predecessors, similar to how a human child learns….
Originally by Jowi Moralesahul of Tom’s Hardware
***
Right off the batch: 50% of Y Combinator 2024 startups built with AI
At Y Combinator, the renowned Silicon Valley startup accelerator, half of the most recent batch of participants are developing projects that incorporate artificial intelligence (AI), according to CEO Garry Tan. In a post announcing the second post-COVID cohort, which gathered in person in San Francisco's Dogpatch neighborhood, Tan highlighted the benefits of the program, stating, "Hundreds of founders are meeting daily for office hours and events, bumping into legendary people, and surrounding themselves with people who will go deep on subjects and celebrate hard work."
Lindsay Amos, Y Combinator's director of communication, further emphasized the unique environment at Y Combinator, stating via email, "There is no other place in the world with the same density of great startup founders per square mile as the area around the YC office—and this is fostering innovation and growth."
San Francisco is experiencing a resurgence in AI-related startups, as evidenced by Crunchbase data. Jared Friedman, a group partner at Y Combinator, noted that these companies are often founded by younger entrepreneurs, with 30% of the last batch consisting of college students or recent graduates, compared to just 10% two years ago. Friedman stated, "Because of AI, it’s the best time in a decade for college students to start startups."
The majority of the companies in the batch are focused on enterprise software as a service (SaaS), with more than two-thirds falling into this category. Approximately 11% of the companies are consumer-oriented, with AI serving as the driving force behind many of these….