MIT Researchers Plan to Remove YouTube Buffering with AI Technology

Ah. Buffering. The ultimate test of our patience. You’re watching an intense moment and the video suddenly hangs or the resolution goes kaput. All you can do is get annoyed and curse your internet connection. If this sounds familiar, there’s good news. Researchers at MIT are coming up with a cool plan – using machine learning to solve this problem and help you enjoy maximum quality without the abrupt interruption or the pixelated resolution.

While Netflix and YouTube have done quite a good job in controlling buffer times, they do so at the cost of picture quality. When there’s a slow connection, YouTube makes the resolution lower to make sure you can still watch uninterrupted. But if you’ve ever tried skipping or moving forward your video hangs and starts re-buffering. This essentially means the video has not been able to re-buffer the next clip rate and therefore is not able to move on with the content.

Headed by Professor Mohammad Alizadeh, MIT’s Computer Science and Artificial Intelligence Laboratory developed what they call the ‘Pensieve,’ system. Using machine learning, the system will be detecting various network conditions where streaming and buffering may suffer. For example, networks usually go haywire in tunnel ways and underground platforms; which means that you won’t be able to stream video content when in either of them. The AI will be detecting such environments and will adjust the speed rate of the video according to the environment. In multiple experiments on a huge variety of Netflix content, the team was able to cut down buffer rates in low connection areas by 30%. So for example when a user is about to enter an underground tunnel in a dead zone, YouTube could turn down the bitrate so that it can load the video without having to rebuffer during the loss of network.

The team is hopeful that this technology will be quite beneficial in streaming high-resolution VR content for the future. Passionate about the project, Alizadeh predicts the future with Pensieve as he says, ‘“The bitrates you need for 4K-quality VR can easily top hundreds of megabits per second, which today’s networks simply can’t support,” Alizadeh says. “We’re excited to see what systems like Pensieve can do for things like VR. This is really just the first step in seeing what we can do.”

For now, though, you can enjoy your Netflix and YouTube viewing buffer-free and with perfect resolutions until the paper is presented (and applauded) at the SIGCOMM Conference in Los Angeles.

P.S – if you still have streaming issues and can’t wait so long for a fix, here are a few quick things you can do.

- Invest a little more in a good internet plan – If you’re not happy with your current plan, make the switch. Go for a network that has super fast download as well as upload speeds.

- Probably try using less HD – Unless needed, like for watching Game of Thrones, try using less of HD. According to Netflix, one hour of HD content eats 3GB of data. That’s a lot especially if you have a limited cap plan. Save the HD for something really good. If you’ve ever read Netflix’s recommended speeds then you’d know that you need at least 5Mbps for HD Quality and 25Mbps for 4k quality.

- Pause the video, go make popcorn – We generally prefer the whole video to load before you play it. Note that is advice is for super slow connections (between 1Mbps to 3Mbps) so if you have excellent internet, it’s OK to make the popcorn while leaving the video on. If not, use the pause button and let the video buffer.

- Make sure you don’t have other downloads in the background – If you’re an active torrent user, chances are that you’d probably have other downloads in the background. Clear them all and don’t let other data hog your streaming data.

If all else fails, clear your cache, restart your Wi-fi device and call for your IP’s customer support.

Farah tries to keep up with the fast-paced tech world by writing about it. She covers latest tech news and writes informative pieces to help her readers make informed decisions about their tech preferences.