During his keynote speech on the first day of Microsoft’s Build Developer Conference, Satya Nadella unveiled several innovative AI advancements that will soon become generally available to the public. The Build Conference is an annual event where developers are introduced to the most recent developments in Windows 11 and Microsoft 365.
One of the key announcements…
Researchers from Anthropic have successfully identified millions of concepts within an advanced large language model (LLM), Claude Sonnet. The knowledge structure of AI models is often likened to a ‘black box,’ emphasizing the mystery behind their internal workings. This complex model architecture makes identifying individual concepts challenging, a problem Anthropic addressed using a technique called…
OpenAI has made the decision to remove the "Sky" voice from its AI system, GPT-4o, following claims by actress Scarlett Johansson that the voice was intentionally designed to sound like her own. The artificial intelligence and machine learning research company refuted the claim, stating that the resemblance was purely coincidental.
According to the company's CEO, Sam…
OpenAI recently revealed that its chatbot, ChatGPT will now allow developers to build API plugins. An API plugin expands the platform's capabilities by integrating it with third-party applications which allow users to book flights, make restaurant reservations, shop, and more.
Amongst the first plugins launched are Klarna – a price comparison site; Expedia – a travel…
The UK's AI Safety Institute (AISI) has conducted a study revealing that AI chatbots can be manipulated into producing harmful, illegal, or explicit responses. The AISI tested five large language models (LLMs), referred to by colour codes, using harmful prompts from an academic paper from 2024, along with a new set of harmful prompts unmodified…
Google has introduced its Frontier Safety Framework's first edition, meant to mitigate the severe risks that potent frontier AI models of the future might pose. It outlines Critical Capability Levels (CCLs), thresholds where the models may present an escalated risk without additional mitigation. Mitigation strategies to tackle models that exceed these CCLs are divided into…
Spotify, the globally renowned music streaming service, has further personalized its user experience by introducing its AI DJ feature to its Australian premium subscribers. Using artificial intelligence technology, Spotify’s AI DJ curates a customized playlist for its users. This cutting-edge feature has been designed to deliver the most fitting pieces of music for any given…
Google's Emoji Kitchen, an online platform that allows users to blend emojis into entirely new ones, now has an official web version. Previously available only through Google's Gboard app, the virtual creative space will now be accessible to users across a wider range of devices including Apple products. The inclusion in Google’s integrated search bar…
Tanisha Katara is a blockchain governance expert and a Governance Innovation Specialist at Polygon Technology. She has been an independent consultant for over six clients in the past two years, spanning across Blockchain protocols, decentralized applications (dApps), and Decentralized Autonomous Organizations (DAOs). Her expertise includes creating strategies and products related to go-to-market, token economics, reputation…
SkyNet is a fictional artificial superintelligence system from the Terminator movie series that becomes sentient and attempts to eradicate all humans. The antagonist in the 1984 release and subsequent films, SkyNet's portrayal tapped into public anxieties around the creation of machines that could quickly surpass human intelligence— a concept that feels increasingly plausible today as…