We consider linear prediction with a convex Lipschitz loss, or more generally, stochastic convex optimization problems of generalized linear form, i.e.~where each instantaneous loss is a scalar convex function of a linear function. We show that in this setting, early stopped Gradient Descent (GD), without any explicit regularization or projection, ensures excess error at most $\varepsilon$ (compared to the best possible with unit Euclidean norm) with an optimal, up to logarithmic factors, sample complexity of $\tilde{O}(1/\varepsilon^2)$ and only $\tilde{O}(1/\varepsilon^2)$ iterations. This contrasts with general stochastic convex optimization, where $\Omega(1/\varepsilon^4)$ iterations are needed Amir et al. 2021. The lower iteration complexity is ensured by leveraging uniform convergence rather than stability. But instead of uniform convergence in a norm ball, which we show can guarantee suboptimal learning using $\Theta(1/\varepsilon^4)$ samples, we rely on uniform convergence in a distribution-dependent ball.

Now we are getting to the biggest and weirdest risk of AI: a super intelligence emerging and wiping out humanity in pursuit of its own goals. To a lot of people this seems like a totally absurd idea, held only by a tiny fringe of people who appear weird and borderline culty. It seems so far out there and also so huge that most people wind up dismissing it and/or forgetting about shortly after hearing it. There is a big similarity here to the climate crisis, where the more extreme views are widely dismissed. In case you have not encountered the argument yet, let me give a very brief summary (Nick Bostrom has an entire book on the topic and Eliezer Yudkowsky has been blogging about it for two decades, so this will be super compressed by comparison): A superintelligence when it emerges will be pursuing its own set of goals.

The power of AI – providing simple solutions to complex business problems Fountech design, develop and integrate AI into the core of your business, often by releasing the untapped potential of Big Data. Sometimes that's data you'll already have, sometimes we'll enable you to find it. We regard ourselves as an AI think-tank, rather than just a development company. Our approach can turn your business ideas into tangible results using targeted AI. That's why our core philosophy is: 'you don't just learn Artificial Intelligence - you need to think it'.

The industrial revolution allowed us to build products at faster rates we had ever seen, and allowed us to scale up our creations to sizes never possible before. Just like machines whose strength is hundreds, if not thousands of times stronger than us, AI's intelligence will be hundreds, if not thousands of times smarter than us. Physical problems like will be solved thousands of times faster than humans could. Machines removed the physical constraints of humans and freed us to pursue more intellectual paths like the information industry, and AI will remove our mental constraints.

How does this lead to innovation? What does this have to do with Deep Learning? Intuition like consciousness is something that we are all aware of its existence but likely have not investigated in enough detail to have a grounded understanding of its nature. In fact, I would say that there's more research on the nature of consciousness than research on intuition. I've written earlier about a few research groups that have explored consciousness with respect to an artificial general intelligence, however I don't think has been equivalently the same effort with the study of intuition.

A new underwater robot is targeting the stunning but dangerous lionfish, which has spread over the Caribbean, the Gulf of Mexico and up the U.S. East Coast as far north as New York's Long Island, with its venomous spines that are dangerous to touch. The robot is the creation of Colin Angle, chief executive officer of IRobot, which makes the Roomba vacuum cleaner. Along with his wife, Erika, and colleagues, he created a new nonprofit to turn automation into environmental tools.

In first generation languages like FORTRAN and C, the burden was on programmers to translate high-level concepts into code. With modern programming languages--I'll use Python as an example--we use functions, objects, modules, and libraries to extend the language, and that doesn't just make programs better, it changes what programming is. Programming used to be about translation: expressing ideas in natural language, working with them in math notation, then writing flowcharts and pseudocode, and finally writing a program. Translation was necessary because each language offers different capabilities. Natural language is expressive and readable, pseudocode is more precise, math notation is concise, and code is executable. But the price of translation is that we are limited to the subset of ideas we can express effectively in each language.

When was the last time you asked your computer something? There's Siri, Google, and Cortana of course, but these systems, clever as they may be, are the thin end of a newly emerging wedge of remarkable new approaches to computer learning and marketing. If you need proof that we are entering a new era of machine learning and artificial intelligence (AI) you need to look no further than Google's DeepMind project. Early this year DeepMind, Google's AI computer, developed initially in London, challenged and beat South Korean Grandmaster Lee Sedol at the ancient game of Go. Why this challenge is so important requires you to think back to the Deep Blue computer, which finally beat Gary Kasparov at chess in the 1990s. Deep Blue had it easy.

We describe an experiment in using IBM’s Watson cognitive system to teach about human-computer co-creativity in a Georgia Tech Spring 2015 class on computational creativity. The project-based class used Watson to support biologically inspired design, a design paradigm that uses biological systems as analogues for inventing technological systems. The twenty-four students in the class self-organized into six teams of four students each, and developed semester-long projects that built on Watson to support biologically inspired design. In this paper, we describe this experiment in using Watson to teach about human-computer co-creativity, present one project in detail, and summarize the remaining five projects. We also draw lessons on building on Watson for (i) supporting biologically inspired design, and (ii) enhancing human-computer co-creativity.

In this paper, we propose a new classification framework for image matrices. The approach is realized by learning two groups of classification vectors for each dimension of the image matrices. One novelty is that we utilize compound regression models in the learning process, which endows the algorithm increased degree of freedom. On top of that, we extend the two-dimensional classification method to a semi-supervised classifier which leverages both labeled and unlabeled data. A fast iterative solution is then proposed to solve the objective function. The proposed method is evaluated by several different applications. The experimental results show that our method outperforms several classification approaches. In addition, we observe that our method attains respectable classification performance even when only few labeled training samples are provided. This advantage is especially desirable for real-world problems since precisely annotated images are scarce.