When we were planning our approach this year to covering Computex, the largest IT trade show in Asia, there was some confusion about where exactly Intel had gone. At that point there was a sense that maybe this year would be a little flat. The Taipei show has always been a big song and dance around the latest CPUs (central processing units) from Intel and the changes they'll bring to computing in the years ahead. As it turned out, Computex was fascinating. On day zero, Nvidia and Asus put on a great show that quickly reminded us that the future is moving beyond the CPU, the chip that traditionally has been the brains of the computer.

Google's cuddly-looking robotic cars have taken a big step on the way to developing a harder edge: they've learned to honk. In the Google Self-Driving Car Project's latest monthly report (PDF), the company says it has been testing horn algorithms in its prototype cars for some time, playing the horn sound inside the cars as a way to make sure they're not beeping in a way that would confuse other drivers. As the algorithm has improved, the cars have recently begun "broadcasting our car horn to the world." Google says its cars are meant to be "polite, considerate, and only honk when it makes driving safer for everyone." That would represent a significant departure from how most humans use their horns.

Already, the electronic brains of the most advanced robotic models surpass human intelligence and are able to do things that will make some of us shudder uncomfortably. But what is your reaction going to be after learning about recent advances in robotics and artificial intelligence? Scientists at the University of Texas (Austin) have simulated mental illness for a computer, testing schizophrenia on artificial intelligence units. The test subject is DISCERN – a supercomputer that functions as a biological neural network and operates using the principles of how human brain functions. In their attempt to recreate the mechanism behind schizophrenia, the scientists have applied the concepts described in the theory of hyper-learning, which states that schizophrenic brain processes and stores too much information too thoroughly by memorizing everything, even the unnecessary details.

If you want to get under Diane Bryant's skin these days, just ask her about GPUs. The head of Intel's data center group was at Computex in Taipei this week, in part to explain how the company's latest Xeon Phi processor is a good fit for machine learning. Machine learning is the process by which companies like Google and Facebook train software to get better at performing AI tasks including computer vision and understanding natural language. It's key to improving all kinds of online services: Google said recently that it's rethinking everything it does around machine learning. "It's a big opportunity, and there will be a hockey stick where every business will be using machine learning," she said in an interview.

The global automotive industry and the oil majors are not known for meekly rolling over when a competitor comes along – from General Motors involvement in killing public transport in Los Angeles in the 1940s to Shell lobbying to undermine EU renewables targets in more recent years. But recently, the world has started to see a new side to the sector: "If you can't beat them, join them; and if you can't join them, buy them out." This week Saudi Arabia announced a surprising new venture for the country's vast oil-generated sovereign investment fund: a 3.5bn stake in the ride-hailing startup Uber. The investment, which values six-year-old Uber at 62.5bn, is one of the largest ever made in a privately held company, and is roughly the same size as the sum total of all investments in the UK's tech sector over the course of 2015, according to the venture capitalist David Galbraith. Saudi Arabia's goal with its investment fund is to use some of the state's 2tn in assets to make long-standing investments that will fund the future of the country once its oil economy begins to sputter.

Nearly 250 million video surveillance cameras have been installed throughout the world, and chances are you've been seen by several of them today. Most people barely notice their presence anymore -- on the streets, inside stores, and even within our homes. We accept the fact that we are constantly being recorded because we expect this to have virtually no impact on our lives. But this balance may soon be upended by advancements in facial recognition technology. Soon anybody with a high-resolution camera and the right software will be able to determine your identity.

Far from the packed arenas and showy lights of the campaign trail, the education of Donald Trump has quietly begun. The unlikely Republican presidential candidate who rose to popularity as a political outsider is now preparing for a general election battle against one of the most experienced policy professionals ever to run for president, presuming Hillary Clinton becomes the Democratic nominee. So as Trump's team begins to ready the candidate for the fall debates and drafts a series of speeches, including one on the U.S. economy, the New York businessman with no elected experience is cramming to get up to speed, consulting experts, asking questions and refining his ideas in anticipation of certain assault by Democrats. The crash courses in foreign and domestic policy are a rite of passage for virtually every presidential candidate. But they carry even greater importance for Trump, who has mocked the Washington establishment as overcomplicating problems and won over many voters with simple ideas that often lack details: build a wall to stop illegal immigration, defeat Islamic State, bring back jobs, make America great again.

This transcript has been automatically generated and may not be 100% accurate.

In a previous post, we've looked at a generative algorithm that can produce images of digits at arbitrary high resolutions, while training on on a set of low resolution images, such as MNIST or CIFAR-10. This post explores several changes to the previous model to produce more interesting results. Specifically, we removed the use of pixel-by-pixel reconstruction loss in the Variation Autoencoder. The discriminator network used to detect fake images is replaced by a classifier network. The generator network used previously had been a relatively large network consisting of 4 layers of 128 fully connected nodes, and we explore replacing this network with a much deeper network of 96 layers, but only with only 6 nodes in each layer.

A Montreal-based platform for sharing information aimed at incoming immigrants called Botler has been launched, developed by recent immigrants to Canada to help facilitate the process for new arrivals to help set up new lives. The platform is referred to as a "next generation" communications tool, using artificial intelligence and machine learning to facilitate group discussion in a hyper-organized and well-designed format. "This idea was born within a community of smart and highly skilled people who dream of a better life. They're willing to take big risks, give up everything and start over," said Botler's founder, Amir Moravej. "Still, so many prospective immigrants are using outdated, poorly designed messaging boards to ask and answer questions about the immigration process."