In-depth articles with more detailed investigations, interesting thought pieces, and more.
Many recently announced “AI agents” are actually sophisticated automation systems, according to Brian Evergreen and Pascal Bornet writing for VentureBeat. True AI agents can research, reason, decide and act to achieve a goal with full autonomy, while automation simply follows predefined recipes when conditions are met. Identifying genuine agency requires looking at a system’s behavior …
Amazon has been working to upgrade its Alexa voice assistant with artificial intelligence to compete with OpenAI’s ChatGPT, but the project has faced repeated delays. According to Austin Carr and Matt Day reporting for Bloomberg, Amazon CEO Andy Jassy tested a prototype in summer 2023 and was initially optimistic, but the planned launch was pushed …
The rise of AI could lead to a 3-12% increase in global electronic waste by 2030, amounting to an extra 2.5 million metric tons annually, according to a study by researchers at the Chinese Academy of Sciences and Reichman University in Israel published in the journal Nature Computational Science. The analysis, based on industry investment …
The research paper “Mind Your Step (by Step): Chain-of-Thought can Reduce Performance on Tasks where Thinking Makes Humans Worse” investigates the effectiveness of chain-of-thought (CoT) prompting in large language and multimodal models. While CoT has generally improved model performance on various tasks, the authors explore scenarios where it may actually hinder performance, drawing parallels from …
I often see discussions about the purpose and benefits of AI tools revolve around whether or not they can completely do your job. The answer is often clear: no, they can’t. If you want, you can always find a way to make AI look stupid. In a way, this is even helpful: you should always …
In 2024, Python overtook JavaScript as the most used programming language on GitHub, according to the platform’s latest Octoverse report. The report, which studies public and open source activity on GitHub, found that AI is rapidly expanding as the global developer community grows. The rise of Python correlates with increasing communities of data scientists, AI …
A small but growing number of workers are becoming “super users” of AI tools like ChatGPT, using them daily to save time and gain skills on the job, according to a Washington Post article by Danielle Abril. These early adopters say AI helps them create content, prepare for meetings, analyze data, build spreadsheets, and even …
A new study by researchers from Technion, Google Research, and Apple reveals that large language models (LLMs) have a deeper understanding of truthfulness than previously thought. The researchers analyzed the internal representations of LLMs across various datasets and found that truthfulness information is concentrated in specific response tokens, VentureBeat reports. By training classifier models on …
Sam Lessin, a general partner at Slow Ventures, revealed on social media that his venture capital firm is successfully using Google’s AI application NotebookLM in place of a traditional customer relationship management (CRM) system. According to an article by VentureBeat’s Carl Franzen, Lessin uploaded 10 years worth of internal emails to NotebookLM and can now …
Meta is well-positioned to benefit from generative AI in digital advertising and content, argues Ben Thompson of Stratechery. Meta’s massive scale, troves of user data, and advanced ad infrastructure enable the company to generate unlimited variations of highly targeted ads and measure their effectiveness, likely boosting ad revenue and margins in the short term. In …