I’ve been playing with local AI models; AI you can run on your own computer for years now. Each year it gets better and better. Similar to the online cloud-based frontier models (you know them). I won’t go into all the advantages of running open source models on your own local hardware, the reasons are somewhat obvious.
In the next decade, AI will likely undergo more significant changes than only becoming more independent; it will also grow more cognitive. AI systems will act as interconnected ecosystems that are capable of contextual awareness, cooperative reasoning, ongoing learning, and adaptive decision-making in almost every facet of society, rather than isolated applications.
Large language models of today are remarkable due to their ability to produce and anticipate information. Persistent memory, multimodal perception, long-term planning, causal reasoning, and self-directed learning within strictly regulated bounds will probably be characteristics of the AI of 2036. Similar to biological neural networks, millions of specialized AI agents will work together to create dynamic intelligence fabrics that continuously optimize national defense, manufacturing, transportation, financial markets, healthcare delivery, and energy grids.
The line between workforce and software will become increasingly hazy. Hundreds of thousands of AI agents working continuously alongside human employees may be employed by organizations as digital workforces. A customized constellation of AI advisers, researchers, legal assistants, financial analysts, engineers, and cybersecurity specialists working around the clock could be present for every knowledge worker. This shift signifies the emergence of an entirely new digital labor force in addition to automation.
Further reading.
Thumbnail image credit: Paper cited in video.
Unsupervised sensory-motor associative learning by human brain explant in-a-dish enables movement imitation by robot.
https://www.researchsquare.com/articl…
The lab.
Not alive. but not dea: disembodied human brains used for drug testing.
https://www.science.org/content/artic…
#science #brain #technology #news #explained
Einstein called imagination “more important than knowledge,” yet we increasingly treat it as a childhood pastime we outgrow. Philosopher of mind Amy Kind argues that it’s something far more practical: the skill we draw on to make our hardest decisions, read the people around us, and work out who we want to become. Like any skill, it weakens without use — and we’re using it less. Reading for pleasure has nearly halved in two decades, fewer parents play with their children each day, and we increasingly hand our creative work to machines. If we don’t make time to exercise it, we’ll lose the capacity to conceive of things being other than they are, and risk being trapped in the present, unable to imagine a different future, let alone build one.
In recent years, amidst the hustle and bustle of contemporary life, people are devoting considerably less time to imaginative activities. Significantly fewer people are reading for pleasure today compared to 20 years ago, and in just the last decade, significantly fewer parents are making time to play with their children on a daily basis than used to be the case.
This neglect of imagination has been accelerated by the increasing reliance on generative AI tools in both personal contexts and professional contexts. In one recent survey, more than 50% of adults reported interacting with AI tool at least several times a week for personal purposes, often for learning, entertainment, or supporting their children’s education. In another study on business uses of AI, more than half of firms surveyed reported using AI in the creation of new products and services and, more generally, in their at innovation projects. With each passing day, we seem to be increasingly more willing—and perhaps even eager—to outsource our creative and imaginative efforts to machines.
A hybrid artificial intelligence model that combines two well-established deep learning techniques has improved the accuracy of financial market forecasts across major stock indices and so-called cryptocurrency, according to work in the International Journal of Reasoning-based Intelligent Systems.
The researchers designed the model, CLSTM-HN, to address a long-standing problem in financial forecasting: balancing the detection of short-term market movements with the recognition of longer-term trends. The researchers tested the system on publicly available data and achieved a forecasting error 15% to 20% lower than that of conventional long-short-term memory (LSTM) models. They also saw an improvement in the accuracy of predicting whether prices would rise or fall by 10% to 14%.
Financial markets are difficult to predict because prices are volatile, noisy and subject to sudden structural shifts. Traditional statistical approaches often rely on assumptions about market behavior that break down during periods of instability.
The explosive growth of data generated by artificial intelligence, cloud computing and modern digital infrastructure is placing increasing pressure on existing information storage technologies. Although magnetic storage systems such as hard disk drives remain the dominant platform for digital storage, they face challenges including rising costs, limited lifespan and relatively slow information retrieval.
To address these challenges, researchers at the University of California, Los Angeles (UCLA) have developed a new optical information storage platform that uses engineered diffractive structures to store and rapidly retrieve thousands of images.
The UCLA team introduced a wavelength-multiplexed diffractive optical storage system composed of multiple passive dielectric layers that are spatially engineered using deep learning. The research is published in the journal Advanced Photonics.
Most contemporary portable electronics, including laptops, smartphones and smart watches, are powered by batteries that need to be recharged daily or every few days. Over the past decade, however, some engineers have been exploring the possibility of developing battery-free electronic devices that autonomously derive electricity from renewable sources, such as sunlight, indoor lighting or heat.
A research team at Penn State University recently developed a compact integrated circuit (IC) that harvests energy solely from ambient light, using this energy to run computations and sense chemicals in its surroundings. This new chip, introduced in a paper published in Nature Electronics, could enable the development of devices that never require charging and thus continue working uninterrupted even in environments where replacement batteries and electrical sockets are not available.
“This work grew out of a broader question we have been asking in my group: Can we build electronic systems that do not simply sense information, but also process that information locally and power themselves from their environment?” Saptarshi Das, senior author of the paper, told Tech Xplore. “Many future Internet of Things (IoT) and edge-computing systems will need to operate in remote or hard-to-access locations, where replacing batteries is impractical. We wanted to demonstrate a compact, fully integrated chip that combines energy harvesting, sensing and computation in a single monolithic three-dimensional architecture.”