Search AI Products and News

A recent joint study by Google, Carnegie Mellon University, and MultiOn explores the application of synthetic data in training large language models. According to Epoch AI, a research institution focused on AI development, currently available high-quality text training data totals around 300 trillion tokens. However, with the rapid advancement of large models like ChatGPT, the demand for training data is growing exponentially, projected to exhaust existing resources by 2026. Therefore, synthetic data is becoming increasingly crucial.

Recently, a survey commissioned by Google Workspace revealed that more and more professionals are willing to utilize artificial intelligence (AI) tools to help them tackle complex and heavy work tasks. Conducted by Harris Poll, the survey covered over 1,000 U.S. employees aged 22 to 39, focusing primarily on participants currently in or aspiring to leadership positions. Image source note: Image generated by AI, image licensed from service provider Midjourney. The results showed that 82% of respondents indicated they already.

Google is leveraging artificial intelligence technology to analyze sound data, aiming to predict and detect health issues early, particularly tuberculosis. In partnership with Indian startup Salcit Technologies, the plan is to integrate this innovative technology into smartphones for easier use in resource-limited areas. This method can assist in diagnosing health conditions through simple audio like cough sounds. Google's ongoing exploration in the health sector, including disease detection through olfactory technology, demonstrates its commitment to enhancing health management capabilities through technological means.

["Sakana AI has announced the successful raising of $30 million in a seed round of financing.", "Investors include prominent firms such as Lux Capital and Khosla Ventures from the U.S., as well as well-known companies like Sony, NTT, and KDDI.", "The company will focus on research and development over the next year or two, exploring AI systems based on smaller 'agent' networks.", "Operating in Tokyo distinguishes Sakana AI from its Silicon Valley competitors, emphasizing the strength of Japan's top AI researchers."]

Google DeepMind's latest research finds that large language models have difficulty autonomously correcting reasoning errors. The study found that models often err when trying to correct their initial responses solely based on their own judgment. The researchers believe that in currently sensitive safety scenarios, self-correction capabilities are essential. Achieving consensus responses through voting can ensure consistency but is not equivalent to self-correction. Further research is needed to improve the self-correction abilities of language models.