Original image

The Science of Implanting False Memories

Original image

By Chris Gayomali

We all have a dubious memory or two that we are convinced is real, even if it never happened. For some the memory is harmless, like the song that played during a first kiss. For others, like courtroom witnesses convinced they saw a suspect only to have their account overturned later by DNA testing, the strange whims of memory can result in real-world disaster.

Which is why a new experiment by neuroscience researchers at MIT's Center for Neural Circuit Genetics is equal parts important and terrifying. Writing in the journal Science, Nobel laureate Susumu Tonegawa outlines how he and his team were able to plant false memories in the brains of mice, tricking them into believing in events that never actually occurred.

The technique they employed is called optogenetics, which allows researchers to manipulate individual brain cells with near-pinpoint precision using a tiny, fiber-optic beam of light. As Ars Technica notes, optogentics "has brought surprising technical advances and changed the way many neuroscientists work."

In this case, Tonegawa suggests that all memories, both real and dreamed, rely on the same basic neural circuitry that can be tampered with. Alok Jha at the Guardian explains:

[Researchers] engineered brain cells in the mouse hippocampus, a part of the brain known to be involved in forming memories, to express the gene for a protein called channelrhodopsin. When cells that contain channelrhodopsin are exposed to blue light, they become activated. The researchers also modified the hippocampus cells so that the channelrhodopsin protein would be produced in whichever brain cells the mouse was using to encode its memory engrams.

In the experiment, Tonegawa's team placed the mice in a chamber and allowed them to explore it. As they did so, relevant memory-encoding brain cells were producing the channelrhodopsin protein. The next day, the same mice were placed in a second chamber and given a small electric shock, to encode a fear response. At the same time, the researchers shone light into the mouse brains to activate their memories of the first chamber. That way, the mice learned to associate fear of the electric shock with the memory of the first chamber. [Guardian]

Then, when researchers placed the mice back in the first chamber, the mice responded in a way that clearly communicated fear: They froze.

"We call this 'incepting' or implanting false memories in a mouse brain," Tonegawa tells Science.

What does that mean for human memory formation? Admittedly, not much for now. At that level of brain activity, "the difference between a mouse and a human are quite small," Edvard I. Moser, a neuroscientist who was not part of the experiment, tells the New York Times. "[But] what I find fascinating about this is that you actually can point to a physical substrate to memory," or an engram. It's like pointing to a specific spot in the brain, and being able to say, "That is the memory."

If the thought of being implanted with a fake memory creeps you out, that's totally understandable. But Tonegawa says his research isn't malicious, Inception-inspired Nolan-ites be damned.

Instead, Tonegawa tells the Timesthe hope is that his advances will illustrate just "how unreliable human memory is." Especially in the courtroom.

More from The Week...

6 Vintage Surfing Photos


The Ultra-Violent Origin of Monogamy


5 of History's Biggest Killjoys

Original image
Google's AI Can Make Its Own AI Now
Original image

Artificial intelligence is advanced enough to do some pretty complicated things: read lips, mimic sounds, analyze photographs of food, and even design beer. Unfortunately, even people who have plenty of coding knowledge might not know how to create the kind of algorithm that can perform these tasks. Google wants to bring the ability to harness artificial intelligence to more people, though, and according to WIRED, it's doing that by teaching machine-learning software to make more machine-learning software.

The project is called AutoML, and it's designed to come up with better machine-learning software than humans can. As algorithms become more important in scientific research, healthcare, and other fields outside the direct scope of robotics and math, the number of people who could benefit from using AI has outstripped the number of people who actually know how to set up a useful machine-learning program. Though computers can do a lot, according to Google, human experts are still needed to do things like preprocess the data, set parameters, and analyze the results. These are tasks that even developers may not have experience in.

The idea behind AutoML is that people who aren't hyper-specialists in the machine-learning field will be able to use AutoML to create their own machine-learning algorithms, without having to do as much legwork. It can also limit the amount of menial labor developers have to do, since the software can do the work of training the resulting neural networks, which often involves a lot of trial and error, as WIRED writes.

Aside from giving robots the ability to turn around and make new robots—somewhere, a novelist is plotting out a dystopian sci-fi story around that idea—it could make machine learning more accessible for people who don't work at Google, too. Companies and academic researchers are already trying to deploy AI to calculate calories based on food photos, find the best way to teach kids, and identify health risks in medical patients. Making it easier to create sophisticated machine-learning programs could lead to even more uses.

[h/t WIRED]

Original image
Land Cover CCI, ESA
Afternoon Map
European Space Agency Releases First High-Res Land Cover Map of Africa
Original image
Land Cover CCI, ESA

This isn’t just any image of Africa. It represents the first of its kind: a high-resolution map of the different types of land cover that are found on the continent, released by The European Space Agency, as Travel + Leisure reports.

Land cover maps depict the different physical materials that cover the Earth, whether that material is vegetation, wetlands, concrete, or sand. They can be used to track the growth of cities, assess flooding, keep tabs on environmental issues like deforestation or desertification, and more.

The newly released land cover map of Africa shows the continent at an extremely detailed resolution. Each pixel represents just 65.6 feet (20 meters) on the ground. It’s designed to help researchers model the extent of climate change across Africa, study biodiversity and natural resources, and see how land use is changing, among other applications.

Developed as part of the Climate Change Initiative (CCI) Land Cover project, the space agency gathered a full year’s worth of data from its Sentinel-2A satellite to create the map. In total, the image is made from 90 terabytes of data—180,000 images—taken between December 2015 and December 2016.

The map is so large and detailed that the space agency created its own online viewer for it. You can dive further into the image here.

And keep watch: A better map might be close at hand. In March, the ESA launched the Sentinal-2B satellite, which it says will make a global map at a 32.8 feet-per-pixel (10 meters) resolution possible.

[h/t Travel + Leisure]


More from mental floss studios