Tag Archives: Computer Science

The Mathematically Perfect Couch

Anyone who has participated in the urban ritual of moving big things through small doorways understands the unique struggle of moving sofa couches. Luckily, mathematicians have found the answer to the woes of furniture movers.

“No John, twist it MY clockwise but push to YOUR left.” | Philip Lee Harvey, Stone, Getty Images

This problem was first formalized in the 1960’s by Leo Moser: what is the largest sofa that can fit around a hallway corner?

Of course, an experienced mover will tell you to stand the sofa on one end, but in what came to be known as the moving sofa problem, we imagine a really, really heavy sofa that is impossible to lift, tilt, or even squish.

Though the problem is simple to understand, it has remained unsolved for over half a century.

First, mathematicians realized the simplest shape to get around a corner doesn’t need to be rotated: a square. If we think of each side of this square as being 1, it has the area=1.

Next, mathematicians understood they could use rotation to help solve the problem and showed a half-circle with an area of about 1.5 would squeeze through the corner.

Square Sofa | Dan Romik

Half-circle Sofa | Dan Romik

By combining these shapes, John Hammersley designed a sofa in 1968

 

That random couch in your grandparents’ basement | Ronald Crufke, August 2010 Ugliest Couch, Norwood Mall

Hideous. No, the Hammersley Couch actually looks like this, with an area of  2.2, more than double the square sofa! He considered this solution to be the best possible.

Hammersley Sofa | Dan Romik

However, in the 1990’s the mathematician Joseph Gerver dropped a bombshell. He toppled Hammersley’s record sofa with complex mathematics, creating a truly marvellous eighteen sided couch. Prepare yourself, this is the most beautiful sofa thus-known to humankind.

The 18-sided Gerver Sofa | Weburbia, Wikimedia Commons

Okay, that may have been an exaggeration. By slightly modifying Hammersley’s Couch design, Gerver was able to increase the area by a whopping 0.5%.

But in mathematics, only one example of something contrary to the rule is enough to disprove the previous finding.

But there’s a common problem with all of these couches – they only turn one direction! What happens if you live somewhere that has both left and right 90° turns?

Unfortunately for Ross, his friends were not mathematicians that could help him derive the optimum couch shape for optimum pivotability | tenor

Enter Dan Romik, a prolific mathematician at the University of California, Davis. Using similar techniques to Gerver, he recently found the likely optimum shape for this unique version of the moving sofa problem.

Romik’s ‘ambidextrous’ sofa | Dan Romik

Romik’s results not only look like a cool modern design for two chairs attached by a table, they also led to surprisingly simple solutions to complex mathematical problems. However, there are still many questions left open in his paper – no one has proved the optimum shape.

Often with difficult mathematical problems, new fields must be developed in order to solve them definitively. There is still much left unproven in mathematics.

– Braydan Pastucha

Cloud Computing: High Performance Computing on any Device?

Can you imagine a smart phone that plays the latest PC games? NVIDIA, a graphics cards company, recently made it possible using a platform called GRID. GRID uses cloud computing technology that allows its users to run accelerated desktops on their phones and laptops. By using these desktops, people can play games and run software with an enhanced speed, because the cloud handles the processing part.

cloud computing. source: ibm.com

For example, when users start GRID on their smart phones, the selected games or software will be operated on cloud servers; these cloud servers are far away from the users but high-performance. On the same time, the real-time display of the games/software will be transferred from the cloud servers to their phones via the Internet. The real-time inputs, such us keyboard and mouse inputs, will be transferred from the phones to the servers as well. This is not only how GRID works but also how cloud computing works.

cloud computing: gaming everywhere on any device. source: nvidia.co.uk

As a result, the users feel like the game/software is running on their phones, and the phones perform well. In fact, the mobile devices do not need to actually process since all the processing is done in the cloud. Similarly, users can use GRID to run any software on any device with an Internet access.

high-performance mobile devices are usually heavy and large.                               source: nextgengamingblog.com

In past decades, IT companies and scientists were trying to make mobile devices high-performance, especially when dealing with graphics. However, poor heat radiating and poor battery life are still big problems for mobile devices due to their small physical size. NVIDIA was facing the same problem, but they took a completely different approach to achieve the same goal. In terms of performance and convenience, they succeed.

cloud computing: work everywhere on any device. source: http://www.nvidia.com

NVIDIA GRID, and other similar services, gives small companies and low-income people an alternative way to use high-performance devices with an affordable price. For instance, a small company can save money on buying high-end PCs, and low-income people can play games on their low-priced laptops.

However, GRID still has the same lag problems as most online games have. In addition, users need to pay for this service by month or by year. Consequently, playing games and working with a desktop computer is still the best plan. However, if a person travels a lot, using GRID can be a good plan for him/her.

 

By Max Ma.

Artificial Intelligence: A Friend or Enemy of Humanity?

During the most recent decade, there have been a lot of movies showing humans fighting against robots. Most of these movies have a background that the robots massively destroying human society, such as The Matrix and The Terminator series. This makes some people worry about the future of human society: Will scientists eventually develop an artificially intelligent machine that can outsmart human in every aspect, and will this machine likely declare a war to overthrow the domination of humans?  Unfortunately, according to some leading scientists such as Stephen Hawking, both of these worries have a high possibility to occur soon.

The revolution? — From Wikimedia Commons

Artificial intelligence (AI), which is defined as the machines or software that are designed to perform tasks, is extremely beneficial to the current society. Therefore, techniques that create advanced AI are developing at a staggering rate nowadays. Some AIs that exist today can even outsmart humans in certain aspects since AIs’ components provide AIs a processing rate that no human being can reach. For example, a well-known AI that is designed to play Go (encirclement chess) defeated the top Go player in 2016 because this AI can always calculate the move that has the highest win chance. Moreover, developers have developed a machine that has the functionality to self-code in order to generate new algorithms for new questions. Using this technology, with enough time and memory space, building an AI that could learn more than any people and outsmart humans in every aspect will become highly possible. In 2013, a research survey asked hundreds of AI experts about the approximate time to achieve a human-level AI. The results were that with the current trends of the development of artificial intelligence, AI experts expect the first human-level AI to be developed in 2022 at the earliest.

What is likely to be the consequence of this development? Although it sounds impossible that a piece of code or a metal machine will declare a war on human beings, a lot of professors have expressed their worries about that a high-level AI, such as a human-level AI, may declare a war on its creators, human beings. The following video shows the details about their worries:

https://youtu.be/LjWc2vtbn9M

To sum up, a genuinely smart AI is likely to be invented by humans in the coming future as technology develops, and such an AI will have the high potential to spell the end of humanity. However, this should not be a reason for stopping the development of AI, as AI has brought so many benefits to humans. From my perspective, as long as the developers have the overall control of any of its developments, such as inserting a self-destruct function triggered by anti-human thoughts for every AI, humans will never lose control to AIs.

-Zhaolin Deng