Page 18 – Advance Aii

#AAAI2025 workshops round-up 3: Neural reasoning...

Images from the workshop on “Neural Reasoning and Mathematical Discovery – An Interdisciplinary Two-Way Street”. In this series of articles, we’re publishing summaries with some of the key takeaways from a few of the workshops held at the 39th Annual AAAI Conference on Artificial Intelligence (AAAI 2025). In this third round-up article, we hear from the organisers of the workshops on: Neural Reasoning and Mathematical Discovery – An Interdisciplinary Two-Way Street AI to Accelerate Science and Engineering Neural Reasoning and Mathematical Discovery – An Interdisciplinary Two-Way Street By Tiansi Dong Organisers: Challenger Mishra, Mateja Jamnik, Pietro Liò, Tiansi Dong. Recent progress in Sphere Neural Networks demonstrates various possibilities for neural networks to achieve symbolic-level reasoning. This workshop aimed to reconsider various problems and discuss walk-round solutions in the two-way street commingling of neural networks and mathematics. Some key takeaways from the workshop were as follows: Black-box neural networks can be successfully used to automatically raise mathematical conjectures and identities and generate new geometries. Irrelevant to the amount of training data, black-box neural networks cannot reach symbolic-level logical reasoning. Interdisciplinary approaches, from philosophy and neuroscience to mathematical modelling and artificial neural networks, can be successfully applied to scientific research, such as...

Why do we – as a human civilization – need to nurture scientific competencies and foster R&D-driven innovation? Can’t conventional techniques and approaches be followed for eternity? Well, the very purpose of science and technology is to uplift humans, elevate lifestyles, and ultimately make the world a better place. Specifically, in the realm of healthcare, scientific advancements are what helps us evolve into smarter and healthier species in the visions of Darwin. And right now, we are at the cusp of such a transformative era. This is the age of Artificial Intelligence (AI) and its myriad applications and use cases such as Large Language Models in healthcare. With the use of such technology, we are closer to solving age-old mysteries relating to the human body, discovering drugs to treat terminal illnesses, and even defying aging. So, buckle up for an interesting article today as we explore the role of LLMs in clinical applications, and how it enables scientific evolution. Interesting Statistics On AI In Healthcare The adoption of AI in healthcare is rapidly accelerating, with tangible results that highlight its transformative impact: 20% reduction in time spent on redundant administrative tasks through AI-powered automation. Over 90% of hospitals are expected...

object(WP_Post)#8435 (24) { [“ID”]=> int(40160) [“post_author”]=> string(2) “36” [“post_date”]=> string(19) “2025-05-26 02:54:37” [“post_date_gmt”]=> string(19) “2025-05-26 02:54:37” [“post_content”]=> string(3706) “ Across the communications landscape, teams are being asked to do more with less, while staying aligned, responsive and compliant in the face of complex and often shifting stakeholder demands. In that environment, how we track, report and manage our relationships really matters. In too many organisations, relationship management is still built around tools designed for customer sales. CRM systems, built for structured pipelines and linear user journeys, have long been the default for managing contact databases. They work well for sales and customer service functions. But for communications professionals managing journalists, political offices, internal leaders and external advocates, these tools often fall short. Stakeholder relationships don’t follow a straight line. They change depending on context, shaped by policy shifts, public sentiment, media narratives or crisis response. A stakeholder may be supportive one week and critical the next. They often hold more than one role, and their influence doesn’t fit neatly into a funnel or metric. Managing these relationships requires more than contact management. It requires context. The ability to see not just who you spoke to, but why, and what happened...

When it comes to real-time AI-driven applications like self-driving cars or healthcare monitoring, even an extra second to process an input could have serious consequences. Real-time AI applications require reliable GPUs and processing power, which has been very expensive and cost-prohibitive for many applications – until now. By adopting an optimizing inference process, businesses can not only maximize AI efficiency; they can also reduce energy consumption and operational costs (by up to 90%); enhance privacy and security; and even improve customer satisfaction. Common inference issues Some of the most common issues faced by companies when it comes to managing AI efficiencies include underutilized GPU clusters, default to general purpose models and lack of insight into associated costs. Teams often provision GPU clusters for peak load, but between 70 and 80 percent of the time, they’re underutilized due to uneven workflows. Additionally, teams default to large general-purpose models (GPT-4, Claude) even for tasks that could run on smaller, cheaper open-source models. The reasons? A lack of knowledge and a steep learning curve with building custom models. Finally, engineers typically lack insight into the real-time cost for each request, leading to hefty bills. Tools like PromptLayer, Helicone can help to provide this...

About a year ago we incorporated a new type of feature into one of our models used for recommending content items to our users. I’m talking about the thumbnail of the content item: Up until that point we used the item’s title and metadata features. The title is easier to work with compared to the thumbnail — machine learning wise. Our model has matured and it was time to add the thumbnail to the party. This decision was the first step towards a horrible bias introduced into our train-test split procedure. Let me unfold the story… Setting the scene From our experience it’s hard to incorporate multiple types of features into a unified model. So we decided to take baby steps, and add the thumbnail to a model that uses only one feature — the title. There’s one thing you need to take into account when working with these two features, and that’s data leakage. When working with the title only, you can naively split your dataset into train-test randomly — after removing items with the same title. However, you can’t apply random split when you work with both the title and the thumbnail. That’s because many items share the...

When most people think of AI, they picture futuristic technology taking over decision-making processes. But according to Jared Peterson, VP of Platform Engineering at SAS, the real value of AI isn’t replacing humans – it’s changing the way you work and run your organizations. Peterson’s presentation at SAS Innovate 2025 was a deep dive into how SAS is pushing the boundaries of productivity, innovation and AI to help organizations be successful. Through captivating demos and inspiring stories, he illustrated the powerful new tools and technologies that are reshaping industries. Here are five key takeaways that highlight the future of AI, data and innovation at SAS. 1. Innovation starts with real developer needs Innovation often comes from understanding real-world frustrations and it’s not always top-down. Peterson shared a powerful story about how SAS® Viya® Workbench came to be. The tool started as a developer’s solution for quickly scaling cloud-native computing. “Sometimes the best ideas don’t come from the top,” Peterson said, explaining that developers on the ground often spot solutions before anyone else does. A moment of frustration led to a breakthrough and that’s how SAS Viya Workbench evolved into what it is today. “Innovation and productivity go hand-in-hand.”— Jared Peterson...

January 15, 2017 Vasilis Vryniotis . 1 Comment Datumbox Framework v0.8.0 is out and packs several powerful features! This version brings new Preprocessing, Feature Selection and Model Selection algorithms, new powerful Storage Engines that give better control on how the Models and the Dataframes are saved/loaded, several pre-trained Machine Learning models and lots of memory & speed improvements. Download it now from Github or Maven Central Repository. One of the main targets of version 0.8.0 was to improve the Storage mechanisms of the framework and make disk-based training available to all the supported algorithms. The new storage engines give better control over how and when the models are being persisted. One important change is that the models are not being stored automatically after the fit() method is finished but instead one needs to explicitly call the save() method providing the name of the model. This enables us not only to discard easier temporary algorithms without going through a serialization phase but also to save/load the Dataframes: Configuration configuration = Configuration.getConfiguration(); Dataframe data = ...; //load a dataframe here MaximumEntropy.TrainingParameters params = new MaximumEntropy.TrainingParameters(); MaximumEntropy model = MLBuilder.create(params, getConfiguration()); model.fit(data); model.save("MyModel"); //save the model using the specific name model.close(); ...

There’s a strange loop taking over social media right now. Scroll through TikTok, YouTube Live, or Instagram, and you’ll see a parade of “digital marketing experts” promoting their latest PDF guide, online course, or coaching program. What’s it about? Digital marketing. But not the kind that helps actual businesses improve performance, it’s a course on how to sell a course about selling courses. Welcome to the infinite funnel. Digital Marketing Isn’t New Some of these influencers act like they’ve discovered a goldmine no one else knows about. They pitch digital marketing as a revolutionary idea in 2025, positioning themselves as hustlers in a fresh, untapped niche. What they don’t realize (or ignore) is that digital marketing has been around for decades. The platforms evolve, but the fundamentals … value creation, targeting, conversion … haven’t changed. These self-declared innovators aren’t breaking new ground. They’re selling reheated versions of what thousands of others have already given away for free. What They’re Really Selling Look closer, and the reality becomes obvious: their primary product is a course about how to sell digital marketing courses. It’s a pyramid of PDFs. There’s no end-user value, no client service, no business utility. The only people buying...

It is hard to overstate the impact of this American election; it will change everything, everywhere. From the functioning of democracy within the United States to its influence and perception abroad, the effects will be profound. People in Utah, Ukraine, Uruguay, and even Uganda will feel its repercussions. There is no place on the planet that will remain untouched in one way or another. As Mark Twain once noted, “Truth is stranger than fiction, but it is because Fiction is obliged to stick to possibilities; Truth isn’t.” Yet, at the individual level, things remain unchanged. No matter who is in power, our duty stays the same: to be good people—citizens, parents, friends, neighbors, and professionals. We must fulfill our responsibilities as decent human beings and resist giving in to fear and cynicism. Living with virtue, truth, justice, and compassion is essential. We should strive to be positive role models and make a small difference every day, as much as we can, without excuse. The only things we control are the stories we choose to believe and the actions we take. We cannot control the rest, so it is not worth wasting time or troubling ourselves with what others do or...

AI comes in many different shapes and sizes. That applies to the use cases, the underlying technologies as well as the approaches to adopting AI in your organization. As many organizations are looking to adopt AI, an increasing need for tangible frameworks to understand the technology in a business perspective is requested by leaders in all industries. Some of the key questions asked by leaders are simple. How much time and money is required to adopt AI and solve business problems via AI and what returns do we get for those efforts? That is more than reasonable questions but answering these questions have been an issue in two parts. Firstly the answers have been a moving target with the technology being in an exponential development and as a result the answers of yesterday seem antique today. Secondly the intangible and explorative nature of AI has made it hard to provide such answers at all. But as AI has matured as a technology, and been packaged into products and ready-to-use solutions, these questions are ready to be answered. The products and solutions might come in different levels of abstractions but they are nevertheless ready for being applied to business...