Anthropic has a fast new AI model — and a clever new way to interact with chatbots (4 minute read)
Anthropic's newest model, Claude 3.5 Sonnet, is reportedly equal to or better than OpenAI's GPT-4o and Google's Gemini across a wide variety of tasks. The model performs well in benchmarks, suggesting that Anthropic has built a legitimate competitor in the space. Claude 3.5 Sonnet is available now to web and iOS users and will be available to developers as well. Anthropic is introducing a new feature called Artifacts that will allow users to see, interact with, and edit the results of their Claude requests.
es-toolkit is a JavaScript utility library with a small bundle size and strong type annotations. It offers a variety of everyday utility functions with modern implementations. es-toolkit is designed with performance in mind and reduces JavaScript code by up to 97% compared to other libraries. It has built-in TypeScript support with straightforward types and useful type guards.
New York bans ‘addictive feeds' for teens (3 minute read)
New York has introduced two new bills - one will require parental consent for social media companies to use addictive feeds powered by recommendation algorithms on people under 18 and the other would limit data collection on minors without consent and restrict the sale of such information.
OpenAI's GPT-5 Pushed Back To Late 2025, But Promises Ph.D.-Level Abilities (2 minute read)
GPT-5 will be significantly more intelligent than GPT-4. It is still at least a year and a half from release. The model will have advanced memory and reasoning capabilities and Ph.D.-level intelligence in specific tasks. A short video of OpenAI's Chief Technology Officer Mira Murati comparing the different GPTs is available in the article.
AI Doesn't Kill Jobs? Tell That to Freelancers (8 minute read)
Freelance jobs that require basic writing, coding, or translation are disappearing across job boards. The number of freelance jobs posted on online platforms in the areas in which generative AI excels has dropped by as much as 21% since the debut of ChatGPT and other generative AI models. Some companies are returning to freelancers as they are finding that AI-generated content isn't up to par - while AI-generated material may look good to those in marketing, it often reads as word salad to more knowledgeable people, especially on niche topics.
Google is killing infinite scroll on search results (1 minute read)
Continuous scrolling on Google search results, which was originally introduced in October 2021, has been removed from desktop search results and will be removed from mobile results in the coming months. It has been replaced with Google's classic pagination bar, which allows users to jump to a specific page of search results or simply click 'Next' to see the next page. A 'More results' button will be shown at the bottom of searches on mobile. The change will allow Google to serve search results faster.
Researchers upend AI status quo by eliminating matrix multiplication in LLMs (6 minute read)
Researchers claim to have developed a method of running AI models more efficiently that involves eliminating matrix multiplication from the process. A fundamental redesign of the neural network operations that are currently accelerated by GPU chips, the method could have deep implications for the environmental impact and operational costs of AI systems. It challenges the prevailing paradigm that matrix multiplication operations are indispensable for building high-performing language models. The approach may outperform traditional large language models at very large scales, but this has not been tested due to computational constraints.
Polyfill.io is being used to infect websites with malware. Websites running any JavaScript code from the domain are advised to remove it immediately. The site previously offered code that added functionality to older browsers. A Chinese organization bought the domain earlier this year. Some popular CDN providers have created mirrors of the domain so sites can continue to use it without having to load the malicious code.