The human brain sucks at remembering lists. Think about it:
When you go to the grocery store, how many items can you manage before you have
to write them down? Three? Five? For most of us, if there's any more than that,
we're going to get back home and find out we forgot the milk (which by the way
was the whole fucking reason we went to the store in the first place).
That's weird, because there are other things in life we have
no problem with. For instance, we don't have much trouble remembering the
locations of a hundred different spots around town, even if we don't know the
addresses (do you even know the street address of your favorite coffee shop?),
or the locations of a thousand items around the house. Sure, you couldn't write
them all down, but if a friend asks you where they can find a flashlight,
you're probably going to have an answer. If only there was a way to exploit
this strength to overcome the other weakness ...
Neuralink
According to Musk, the main barrier to human-machine
cooperation is communication "bandwidth."
This means that using a touch screen or a keyboard is a slow
way to communicate with a computer. Musk's new venture aims to create a direct
"high-bandwidth" link between the human brain and machines.
What that system would actually look like is not entirely
clear yet. Words like "neural
lace" and "neural dust" have been bandied about, but
all that has really been revealed is a business model. Neuralink has been
registered as a medical research company, and Musk said the firm will produce a
product to help people with severe brain injuries within four years.
This will lay the groundwork for developing BCIs for healthy
people, thus enabling humans to communicate by "consensual
telepathy," which could be ready within five years, Musk said. Some
scientists, particularly those in the neuroscience community, are skeptical
of Musk's ambitious plans.
Facebook
Not to be outdone, just a few weeks after Musk launched
Neuralink, Facebook announced that it is working on a way to let
people type with their thoughts.
The goal is to build a device that would allow people to
"type" up to 100 words per minute, according to Regina Dugan, head of
the company's secretive Building 8 research group. Dugan also suggested that
the device could work as a "brain mouse" for augmented reality (AR),
removing the need to track hand movements to control cursors, The
Verge reported.
Facebook has also been light on the details of its plans.
The company has said it does not think implants are feasible in the long term,
so it's focusing on developing some kind of cap that could track brain activity
noninvasively, most likely using optical imaging.
But this technology doesn't exist yet. So, in the meantime,
Facebook said that, within two years, it
plans to create a prototype medical
implant that would pave the way for future devices.
Kernel
Musk wasn't the first wealthy entrepreneur to dive into the
underdeveloped neurotechnology space. Last August, Bryan Johnson, founder of
the online payments company Braintree, invested $100 million into the startup
called Kernel.
The company's initial goal was to develop a chip that could
record memories and redeliver them to the brain, based on research by Theodore
Berger, a biomedical engineer and neuroscientist at the University of Southern
California. But six months later, the two parted ways due to the long
timescales involved, reported
MIT Technology Review, and the company is now focusing on technology
similar to Neuralink.
Kernel plans to build a flexible platform for recording and
stimulating neurons, with the goal of treating diseases such as depression and
Alzheimer's. But like Musk, Johnson is not afraid to discuss the prospect of
using the technology to augment human abilities and merge with machines.
"There's this huge potential to co-evolve with our
technology," Johnson
told CNBC.
Emotiv
Unlike some other companies in this burgeoning industry,
Emotiv actually makes products — electroencephalography headsets that record
brain activity noninvasively.
The technology is lower fidelity than the kinds of neural
implants other companies, such as Neuralink, are considering, but it is more
established. The company has a research-grade device, called EPOC+, which sells
for $799. But it also produces a more consumer-oriented headset, called
Insight, which retails for $299.
Emotiv also produces a variety of software products that
allow users to visualize their brain activity in 3D; measure their brain
fitness; and even control drones, robots and video games, reported
The Daily Dot. The company was selected to be part of the Disney
Accelerator program in 2015, with the aim of creating
a "wearable for the brain."
DARPA
Although it's not a company itself, the U.S. military's
Defense Advanced Research Projects Agency announced a $60 million program last
year to develop an
implantable neural interface in collaboration with a consortium of
private companies.
The project, which is a part of former President Barack
Obama's BRAIN Initiative, is ambitious. DARPA wants a device that can record 1
million neurons simultaneously and stimulate at least 100,000 neurons in the
brain. DARPA also wants the device to be wireless, the size of a nickel and
ready in four years, which is an incredibly aggressive deadline, according
to MIT Technology Review.
Potential applications include compensating for sight or
hearing problems because the device could feed digital auditory or visual
information directly into the brain. The exact technological approach is unclear
at this stage, but the project has the heft of some major engineering giants,
such as Qualcomm, behind it, Quartz
reported.
Original article on Live
Science.
No comments:
Post a Comment