Despite plenty of misgivings, artificial intelligence--and in particular, generative AI that produces text and images from prompts--continues to be pushed into the hardware and software we use every day. Microsoft has been active in the space, adding AI chatbot capabilities to its Bing search engine earlier this year, and it's now previewing an early version of its new Copilot AI assistant in Windows 11. Copilot has been built to "enhance your creativity and productivity," Microsoft says, and it works in a similar way to Bing's chatbot--capable of coming up with everything from travel advice to an original poem. To get Copilot in Windows 11, make sure you're running the very latest version of the operating system: Head to Windows Update in Settings to check (you might need to turn on the Get the latest updates as soon as they're available toggle switch). By default, you should see a Copilot button on the taskbar, which you can click to launch it (head to Personalization then Taskbar in Settings if you want to change this).

What would America's top landmarks look like, reimagined by some of the most famous and controversial architects that have ever lived? An Instagram account, Imagined Architecture, created a stir with a'reimagined' White House designed by world-famous architects. With architects ranging from modernist genius Anthony Gaudi and British-Iranian'Queen of Curves' Zaha Hadid, Midjourney has reimagined everything from the Chrysler Building to the Statue of Liberty in typically surreal style. San Francisco-based Midjourney is a rival to OpenAI's Dall-E, which is now integrated into its iconic ChatGPT artificial intelligence chatbot. Like ChatGPT, it can be controlled by simple text prompts, and can generate everything from realistic photographs to paintings: it's controlled through the Discord chat app, and available to subscribers from $10 a month.

Elon Musk's new AI company, xAI, will release its chatbot to subscribers of X's $16 per month Premium plan once it exits beta. The system, called Grok, is positioned to be a competitor to OpenAI's ChatGPT and started rolling out to a select group of users this weekend. As soon as it's out of early beta, xAI's Grok system will be available to all X Premium subscribers Musk shared a few screenshots of the conversational AI on X, and confirmed its responses will unfortunately be laden with Musk-type humor. The CEO also further touted its capabilities compared to the competition, tweeting, "Grok has real-time access to info via the X platform, which is a massive advantage over other models." There's no public timeline yet for when it will be out of beta, but Musk said it "will be available to all X Premium subscribers" when it is.

We've been hearing rumblings about Elon Musk's new AI venture, xAI, for months, and now it looks like it's almost here. The Tesla CEO and noted social media guru just took to Twitter/X to proclaim that his AI venture will launch its first model tomorrow, November 4. This is a beta phase, of a sort, as it's being released only to a "select group", though Musk didn't specify as to what went into the selection process. Will it be a random drop or will the AI model be reserved for "VIPs" like, uh, Tucker Carlson, Chaya Raichik or the indefatigable Catturd? Musk is making lofty promises about his AI, announcing that "in some important respects, it is the best that currently exists." It'll be competing with big-time offerings by OpenAI, Google, Meta and numerous others, so we'll see what "important respects" make it the best that currently exists.

Microsoft has one of the largest and most powerful collections of web servers on the planet. But even it might balk at the thought of a billion or so Windows users hitting data- and processor-intensive generative AI services 24-7. So perhaps it's not surprising that some new language in its online services user license agreement says that Microsoft will employ "temporary throttling of Customer's access to the Microsoft Generative AI service" for excessive use. Exactly what constitutes excessive use of generative AI (which allows a user to create text and images based on specific input, as seen with ChatGPT and DALL-E) is not specified. But as anyone who's tried out these tools knows, it's not an instant process and complex strings of text generation or intricate formatting might take several minutes for a remote server to complete.

Most applications of Artificial Intelligence (AI) are designed for a confined and specific task. However, there are many scenarios that call for a more general AI, capable of solving a wide array of tasks without being specifically designed for them. The term General-Purpose Artificial Intelligence Systems (GPAIS) has been defined to refer to these AI systems. To date, the possibility of an Artificial General Intelligence, powerful enough to perform any intellectual task as if it were human, or even improve it, has remained an aspiration, fiction, and considered a risk for our society. Whilst we might still be far from achieving that, GPAIS is a reality and sitting at the forefront of AI research. This work discusses existing definitions for GPAIS and proposes a new definition that allows for a gradual differentiation among types of GPAIS according to their properties and limitations. We distinguish between closed-world and open-world GPAIS, characterising their degree of autonomy and ability based on several factors such as adaptation to new tasks, competence in domains not intentionally trained for, ability to learn from few data, or proactive acknowledgment of their own limitations. We propose a taxonomy of approaches to realise GPAIS, describing research trends such as the use of AI techniques to improve another AI (AI-powered AI) or (single) foundation models. As a prime example, we delve into GenAI, aligning them with the concepts presented in the taxonomy. We explore multi-modality, which involves fusing various types of data sources to expand the capabilities of GPAIS. Through the proposed definition and taxonomy, our aim is to facilitate research collaboration across different areas that are tackling general purpose tasks, as they share many common aspects. Finally, we discuss the state of GPAIS, prospects, societal implications, and the need for regulation and governance.

Bayesian Optimization (BO) is a technique for sample-efficient black-box optimization that employs probabilistic models to identify promising input locations for evaluation. When dealing with composite-structured functions, such as f=g o h, evaluating a specific location x yields observations of both the final outcome f(x) = g(h(x)) as well as the intermediate output(s) h(x). Previous research has shown that integrating information from these intermediate outputs can enhance BO performance substantially. However, existing methods struggle if the outputs h(x) are high-dimensional. Many relevant problems fall into this setting, including in the context of generative AI, molecular design, or robotics. To effectively tackle these challenges, we introduce Joint Composite Latent Space Bayesian Optimization (JoCo), a novel framework that jointly trains neural network encoders and probabilistic models to adaptively compress high-dimensional input and output spaces into manageable latent representations. This enables viable BO on these compressed representations, allowing JoCo to outperform other state-of-the-art methods in high-dimensional BO on a wide variety of simulated and real-world problems.

Previous efforts to support creative problem-solving have included (a) techniques (such as brainstorming and design thinking) to stimulate creative ideas, and (b) software tools to record and share these ideas. Now, generative AI technologies can suggest new ideas that might never have occurred to the users, and users can then select from these ideas or use them to stimulate even more ideas. Here, we describe such a system, Supermind Ideator. The system uses a large language model (GPT 3.5) and adds prompting, fine tuning, and a user interface specifically designed to help people use creative problem-solving techniques. Some of these techniques can be applied to any problem; others are specifically intended to help generate innovative ideas about how to design groups of people and/or computers ("superminds"). We also describe our early experiences with using this system and suggest ways it could be extended to support additional techniques for other specific problem-solving domains.

Over the last few years, large neural generative models, capable of synthesizing semantically rich passages of text or producing complex images, have recently emerged as a popular representation of what has come to be known as ``generative artificial intelligence'' (generative AI). Beyond opening the door to new opportunities as well as challenges for the domain of statistical machine learning, the rising popularity of generative AI brings with it interesting questions for Cognitive Science, which seeks to discover the nature of the processes that underpin minds and brains as well as to understand how such functionality might be acquired and instantianted in biological (or artificial) substrate. With this goal in mind, we argue that a promising research program lies in the crafting of cognitive architectures, a long-standing tradition of the field, cast fundamentally in terms of neuro-mimetic generative building blocks. Concretely, we discuss the COGnitive Neural GENerative system, such an architecture that casts the Common Model of Cognition in terms of Hebbian adaptation operating in service of optimizing a variational free energy functional.

The General Transit Feed Specification (GTFS) standard for publishing transit data is ubiquitous. GTFS being tabular data, with information spread across different files, necessitates specialized tools or packages to retrieve information. Concurrently, the use of Large Language Models(LLMs) for text and information retrieval is growing. The idea of this research is to see if the current widely adopted LLMs (ChatGPT) are able to understand GTFS and retrieve information from GTFS using natural language instructions without explicitly providing information. In this research, we benchmark OpenAI's GPT-3.5-Turbo and GPT-4 LLMs which are the backbone of ChatGPT. ChatGPT demonstrates a reasonable understanding of GTFS by answering 59.7% (GPT-3.5-Turbo) and 73.3% (GPT-4) of our multiple-choice questions (MCQ) correctly. Furthermore, we evaluated the LLMs on information extraction tasks using a filtered GTFS feed containing four routes. We found that program synthesis techniques outperformed zero-shot approaches, achieving up to 93% (90%) accuracy for simple queries and 61% (41%) for complex ones using GPT-4 (GPT-3.5-Turbo).