DeepSeek’s rapid rise in the AI industry has brought both applause for its cost-effective approaches and skepticism regarding its transparency. While its AI assistant soared to the top of app store charts in both the UK and US, questions remain about how accurately DeepSeek represents its training costs, funding sources, and technological achievements. Meanwhile, new announcements—like DeepSeek’s omnimodal Janus-Pro-7B model—demonstrate the company’s ambitions aren’t slowing down. Cost and Funding Controversies Training Costs Under $6 Million?

DeepSeek, a Chinese startup, has made headlines by releasing an AI assistant that quickly rose to the top of both the UK and US Apple App Store charts. This surge in popularity has placed it above established names like ChatGPT, sparking conversations around innovation, cost-effective AI development, and privacy concerns. Key Highlights from DeepSeek’s Approach Budget-Friendly Development DeepSeek’s DeepSeek-V3 model was developed for under $6 million, which is far less than many competitors spend.

OpenAI recently stirred the AI community with news that its o3-mini model will soon be accessible to a broader audience. While official details remain somewhat sparse, we have enough context—and a few recent updates—to get a clearer picture of what o3-mini might offer, how it builds on o1, and what to expect from this new addition to the OpenAI ecosystem. Quick Highlights: New Info as of January 17, 2025 CEO Announcement: Sam Altman confirmed o3-mini is finalised and set for release in the coming weeks.

The Canvas feature in ChatGPT has been turning heads since it first launched, especially among developers who want a quick, interactive way to test and display code. Until recently, though, Canvas was only available on GPT-4o—which not everyone had access to. Now, it’s rolled out to o1, ChatGPT’s more advanced reasoning model, opening the door for a wider group of users to benefit from real-time HTML and React rendering. Let’s break down why this matters and what you can do with it.

If you’ve ever used an AI tool and wondered, “Where’s this information coming from?”, you’re not alone. Tools like ChatGPT, Perplexity, and NotebookLM already offer some form of source referencing, but Anthropic’s new Citations feature for Claude takes things a step further in terms of precision, flexibility, and deep integration. Let’s explore what makes this development so significant—and how it sets a new standard for trustworthy AI. Overview: Grounding AI Responses in Real Sources Anthropic has rolled out a Citations feature for its Claude AI models—currently Claude 3.

Another week, another leap forward in the AI space! This time it’s Perplexity making waves with their brand-new AI Assistant for Android. If you’re wondering what it does, why it matters, and how it compares to what’s already out there, let’s break it down in a straightforward way. What Is Perplexity’s Android Assistant? Perplexity has taken its AI capabilities and wrapped them into a handy new assistant designed specifically for Android devices.

There’s been a lot of buzz around OpenAI’s latest feature, Operator. If you’re wondering what it is, what it can do, and why people are so excited, don’t worry—you’re not alone. At first glance, it might sound a bit like something out of a sci-fi film, but really, it’s a clever step forward in how we use AI to simplify tasks. Let’s break it down. What Is Operator? Simply put, Operator is like having an AI personal assistant that can use the internet on your behalf.

If there’s one thing you can count on in the fast-paced world of tech, it’s that nothing stays the same for long. This past week has been a whirlwind for us—wrapping up our first Generative AI Skills Bootcamp cohort, hosting back-to-back events, and then watching in disbelief as five major AI announcements dropped just hours after we’d updated everyone on the latest news. Let’s dive in! Monday: Wrapping Up Our First Generative AI Skills Bootcamp Ellie and I run the Generative AI Skills Bootcamp together, and on Monday, we watched our first cohort deliver their final business plan presentations.

The tech to build the holodeck A visit to an elderly mother in Germany highlighted the capability of Gaussian splatting, a novel 3D capture technique creating lifelike 3D models with your phone. This technology, developed in 2023 by European scientists, uses millions of fuzzy, translucent blobs to capture the precise color, location, scale, and transparency of objects, enabling photorealistic 3D reconstructions. Major tech firms like Niantic, Google, Snap, and Meta are utilizing this breakthrough for various applications, from mapping and geospatial games to volumetric video capture and metaverse spaces.

This weekend has been nothing short of extraordinary for TikTok users in the United States. Just this morning, on January 19th, TikTok went dark, leaving millions of users scrambling to adapt and countless small businesses in disarray. Yet, in an astonishing turn of events, the app is now being restored following an intervention by President-elect Donald Trump. All of this has happened within 24 hours — and during a weekend, no less.

As of January 19, 2025, TikTok has officially gone dark in the United States. The popular app, known for its viral dance trends, comedy sketches, and unparalleled ability to catapult creators into stardom, has ceased operations in anticipation of an impending federal ban. For millions of Americans, this marks the end of an era — and the start of a scramble to adapt. Why Did TikTok Shut Down in the U.

The UK government plans to roll out a digital ID system for age verification in pubs, clubs, and shops by 2025, marking a major milestone in modernizing identity verification processes. This initiative is part of the government’s broader ‘Plan for Change’ programme, which aims to enhance digital identity infrastructure and drive economic growth. Key Features of the Digital ID System Smartphone-based: Users will be able to prove their age using certified digital identities stored on their phones.

Microsoft has recently integrated its AI-powered assistant, Copilot, into Microsoft 365 Personal and Family subscriptions for consumers. This integration introduces several new functionalities across various applications, accompanied by a subscription price increase of $3 per month in the U.S. Enhanced Features with Copilot Integration Subscribers now have access to Copilot’s AI capabilities within applications such as Word, Excel, PowerPoint, Outlook, OneNote, and Designer. These features include: Content Generation: In Word and PowerPoint, Copilot assists in drafting text and creating presentations based on user prompts, streamlining the content creation process.

Google Signs Deal with AP to Deliver Up-to-Date News through its Gemini AI Chatbot Google has signed a deal with The Associated Press (AP) to enhance its AI chatbot, Gemini, by providing real-time news updates from AP. This initiative marks Google’s first agreement with a news publisher. AP’s Chief Revenue Officer Kristin Heitmann expressed satisfaction with the collaboration, emphasizing the mutual benefit of accurate and timely news distribution. Financial details of the partnership remain undisclosed, and it’s uncertain how Google will present the news content.

Meta’s Generative AI Push Could Bring Big ‘Windfall’ For WhatsApp, Analyst Says Meta Platforms is testing generative AI-powered chatbots within WhatsApp, aiming to enhance customer service and drive e-commerce sales. Analysts project this could generate incremental revenue of $16.6 billion by 2025 and up to $45 billion by 2030. Businesses are charged 2.5 cents per conversation every 24 hours, with potential for price increases. Meta’s stock has responded positively, reflecting investor confidence in this AI initiative.