If you've been interested in upgrading your AI experience with one of the available paid subscriptions, pay attention because here's your chance to score a free year of Perplexity Pro. Perplexity is a free AI chatbot and internet search tool that operates like other AI chatbots. ZDNET senior contributing editor Steven Vaughan-Nichols called it his favorite chatbot, beating out ChatGPT and Copilot. Perplexity Pro gives you several advantages over a free account. These include increased usage of Pro Search, which breaks down queries with multi-step reasoning and programming; the option to choose which AI model you want to use; file analysis, which lets you upload documents for summaries or insights; multimodal capabilities, where you can capture a picture or screenshot and asking questions about it; and image generation from Access Playground v3, OpenAI DALL-E 3, Flux, and Stable Diffusion XL.

Large language models and other forms of generative artificial intelligence are improving steadily at "self-correction," opening up the possibilities for new kinds of work they can do, including "agentic AI," according to the vice president of Anthropic, a leading vendor of AI models. "It's getting very good at self-correction, self-reasoning," said Michael Gerstenhaber, head of API technologies at Anthropic, which makes the Claude family of LLMs that compete with OpenAI's GPT. "Every couple of months we've come out with a new model that has extended what LLMs can do," said Gerstenhaber during an interview Wednesday in New York with Bloomberg Intelligence's Anurag Rana. "The most interesting thing about this industry is that new use cases are unlocked with every model revision." Also: Anthropic's latest AI model can use a computer just like you - mistakes and all The most recent models include task planning, such as how to carry out tasks on a computer as a person would; for example, ordering pizza online.

Anthropic is making its Claude AI a little more useful by enabling it to analyze your documents. Starting today, you can add a file from Google Docs to your Claude AI chat or project to give more context to conversations. You can then ask the chatbot questions about the document, such as asking for a summary of the content, like in the below screenshot. Currently, Google Docs integration for Claude is only available to users on Claude Pro and Claude for Work. For Claude for Work users, you'll need to follow the instructions here to set up your integration.

Apple's promised Siri overhaul likely won't be fully realized until 2026. A new report from Bloomberg's Mark Gurman says Apple is working on a Siri LLM to compete with ChatGPT. But it won't be ready until the iOS 19 and macOS 16 software updates, likely announced in 2025 and released in spring 2026. The tech giant has been scrambling to update its dated voice assistant in the generative AI era, which has seen the rise of ChatGPT and competitors like Google Gemini. Apple announced a slew of new AI features as part of Apple Intelligence at WWDC last spring, including an updated Siri.

OpenAI and Google, two major players in artificial intelligence (AI), continue to fight for model supremacy in the public forum -- and the race is rapidly heating up. Just a day after OpenAI secured the number one chatbot title in the Chatbot Arena with its GPT-4o update, Google released Gemini Exp 1121, an experimental model that quickly rose to tie with ChatGPT for the top spot. The latest development was announced via an X post on Thursday by the Chatbot Arena's official account, which noted that large language model (LLM) "progress is now measured in days." Woah, huge news again from Chatbot Arena @GoogleDeepMind's just released Gemini (Exp 1121) is back stronger ( 20 points), tied #1 Overall with the latest GPT-4o-1120 in Arena! Compared with Gemini's last release, Exp 1114, the model's update helped it climb from third to first place overall, and specifically improved in hard prompts, coding, math, and creative writing.

Apple may be late to the generative AI party, but that doesn't mean it's standing still. The company is developing a new version of Siri that would work more like popular AI bots such as ChatGPT and Google Gemini, according to Bloomberg's Mark Gurman. The goal is to catch up with other AIs by equipping Siri with more advanced and conversational skills, Gurman said in a story published Thursday, citing "people with knowledge of the matter." Also: 2 reasons why I recommend the M4 Mac Mini to everyone - and it's 100 off right now Dubbed "LLM Siri" by Apple employees (LLM standing for large language model), the new Siri is already being tested internally on iPhones, iPads, and Macs as a standalone app. But if all goes well, the new assistant will eventually replace the Siri we all know and have a hard time loving. Apple could announce the new AI as soon as 2025 as part of iOS 19 and macOS 16.

Did you know ChatGPT can help you find your celebrity look-alike? With just a few selfies, this AI tool can pinpoint your famous doppelgänger in seconds. When I attempted to uncover my A-list twin through the standard ChatGPT window, I was met with this disappointing response: "I cannot determine your celebrity look-a-like from the photo." Fortunately, there is a workaround, allowing you to use ChatGPT as your own personal celebrity look-a-like matching tool. These step-by-step instructions are based on the ChatGPT desktop app, but they should work for the mobile platform as well.

The artificial intelligence startup Anthropic said on Friday it had raised an additional 4bn investment from its longtime backer Amazon.com, Amazon will maintain its position as a minority investor, the company said. Its AWS unit will also be Anthropic's official cloud provider. Anthropic also said it was working with AWS's Annapurna Labs on the development of future generations of Amazon's Trainium chips and plans to train its foundational models on the hardware. Britain's competition regulator had said in September that Amazon's partnership with Anthropic will not be referred for a deeper inquiry as it did not fall under its jurisdiction.

When better to hold a conference on artificial intelligence and the countless ways it is advancing science than in those brief days between the first Nobel prizes being awarded in the field and the winners heading to Stockholm for the lavish white tie ceremony? It was fortuitous timing for Google DeepMind and the Royal Society who this week convened the AI for Science Forum in London. Last month, Google DeepMind bagged the Nobel prize in chemistry a day after AI took the physics prize. Scientists have worked with AI for years, but the latest generation of algorithms have brought us to brink of transformation, Demis Hassabis, the chief executive officer of Google DeepMind, told the meeting. "If we get it right, it should be an incredible new era of discovery and a new golden age, maybe even a kind of new renaissance," he said.

OpenAI has lifted the lid (just a crack) on its safety-testing processes. It has put out two papers describing how it stress-tests its powerful large language models to try to identify potential harmful or otherwise unwanted behavior, an approach known as red-teaming. The first paper describes how OpenAI directs an extensive network of human testers outside the company to vet the behavior of its models before they are released. The second presents a new way to automate parts of the testing process, using a large language model like GPT-4 to come up with novel ways to bypass its own guardrails. MIT Technology Review got an exclusive preview of the work.