Будущее
за
имплантированными мозговыми чипами
Рэй
Курцвейл,
визионер в области искусственного интеллекта, который более 20 лет
назад
относительно точно предсказал нынешнюю хронологию развития
искусственного
интеллекта, недавно в своей книге «Сингулярность ближе» еще раз ясно
дал
понять, что развитие не остановится на искусственном интеллекте.
Будущее за имплантированными мозговыми чипами, так что через несколько
лет мы
сольемся с машинами; знания будут скачиваться, а не изучаться.
Neuralink — предвестник грядущих событий, так сказать, пионер. И он
ясно
показывает, что работает.
Written
by Brooke Becher
Image: Shutterstock
Brain-computer
interfaces (BCI) are devices that create a direct communication pathway
between
a brain’s electrical activity and an external output. Their sensors
capture
electrophysiological signals transmitted between the brain’s neurons
and relay
that information to an external source, like a computer or a robotic
limb,
which essentially lets a person turn their thoughts into actions.
These
brain chips
go over the scalp in a wearable device, get surgically placed under the
scalp or
even get implanted within brain tissue. The idea is that, the closer
the chip
is to the brain’s neural network, the more clear, or “high definition,”
a
signal can be interpreted.
A brain-computer interface (BCI) is a
device that
lets the human brain communicate with and control external software or
hardware, like a computer or robotic limb.
Perhaps the most
popular example of a brain-computer interface is Neuralink,
a chip that was implanted into a quadriplegic patient’s brain in 2024
and
allows him to control a computer.
Brain-computer
interfaces are devices that process brain activity and send signals to
external
software, allowing a user to control devices with their thoughts.
With
BCI
technology, scientists envision a day when patients with paralysis,
muscle
atrophy and other conditions could regain motor functions.
Rehabilitation
services could also adopt BCIs to accelerate recovery from injuries.
Ramses
Alcaide, CEO
of neurotech startup Neurable,
which develops non-invasive brain-computer interfaces in the form of
headphones, sees potential for BCI-enhanced devices to become an
everyday item
for the average person.
“If
we can make
brain-computer interfaces accessible and seamless enough then they can
be
integrated into our daily lives, just as we use smartphones or laptops
today,”
Alcaide told Built In. “But in order to truly become a ubiquitous tool,
they
need to be comfortable, intuitive and reliable enough that people can
use them
without consciously thinking about them — similar to how we use a mouse
or
keyboard to interact with a computer.”
Excitement
around
the possibilities of BCI has resulted in a thriving market, which is expected to triple in
size from
$2 billion in 2023 to $6.2 billion by the end of the decade.
Brain-computer
interfaces are modeled after the electrophysiology of a brain’s neural network. When we make a decision —
or even
think about making a decision — electrical chemical signals spark. This
phenomenon is located in our nervous system; more specifically, in the
gaps
between neurons, known as synapses, as they communicate back and forth.
In
order to capture
this brain activity, BCIs place electrodes proximal to these
conversations.
These sensors detect
voltages, measuring the frequency and intensity of each “spike” as they
fire or
potentially fire.
“It’s
like a
microphone; but in this case, we’re listening to electrical activity
instead of
sound,” said Craig Mermel, president and chief product officer at Precision
Neuroscience, a
startup developing a semi-invasive, reversible neural implant. “We’re
picking
up the electrical chatter of the brain’s neurons communicating with
each
other.”
That
information is
then fed through local computer software, where it’s translated in a
process
known as neural decoding. This is where a variety of machine
learning algorithms
and other artificial intelligence agents take over,
converting
complex data sets collected from brain activity into a programmable
understanding of what the brain’s intention might be.
Brain-computer
interfaces come in two main types: invasive and non-invasive.
“The
near-term goal
[of brain-computer interfaces] is to give the abilities back to those
who have
lost them,” said Sumner Norman, a scientist at nonprofit startup Convergent
Research and
former chief brain-computer interface scientist at software firm AE Studio. “But in the
long term,
this tech is also intended to create a kind of tertiary cortex, or
another
level of the human brain function and an executive function that would
allow us
to be almost superhuman.”
These
are some of
the more common use cases of brain-computer interfaces:
By
supplying a
real-time neural feedback loop that rewires the brain, BCIs are capable
of
restoring movement, mobility and autonomy for paralyzed and disabled
patients,
heightening their quality of life. In more chronic cases, robotic devices and
limbs are integrated.
Headsets
are a way
to deliver a non-invasive approach to brain-computer interfaces.
Some boost productivity and enhance focus, as seen with
Neurable’s Enten, while others restore motor functions to an individual’s
upper
extremities following a stroke, such as the IpsiHand system by
Neurolutions
Inc.
Non-verbal
individuals, who may be stuck in a “locked in” state following a stroke
or
severe injury, can use eye movement for computer-augmented communication.
In several studies,
users have exercised control of social networking apps, email
administration,
virtual assistants and instant message services sans motor skills.
Dimming the
lights or changing the channel on a TV are examples of how BCIs can be
adapted
in the home.
Neuralink, headed by
Elon
Musk, has developed a coin-sized surgical implant. In order to monitor
brain
activity as closely as possible, Neuralink’s device, the Link, uses
micron-scale wires of electrodes that fan out into the brain. Its
primary focus
is to treat paralysis. The company has inserted its implant in one
patient
and has plans to take on a second
patient.
Neurable is
building
headphones that interpret brain signals to level up productivity. Its
first
pair, Enten, uses advanced data
analysis and signal processing techniques to maximize a users’
peak
focus periods throughout the day. The company’s MW75 Neuro builds
on the Enten, offering the same BCI capabilities coupled with more
secure data
encryption and a mobile app that makes it easier to analyze
data-driven insights.
Precision
Neuroscience is
approaching brain-computer interface systems with a surgically
implanted brain
chip that’s minimally invasive and fully reversible. The Layer 7
Cortical
Interface is a thin film of micro-electrodes, about as thick as
one-fifth of a
human hair, that conforms to the brain’s cortex just under the
skull
without damaging any tissue. In June 2023, Precision conducted an in-human clinical study of its
neural
implant. It has since expanded its research to include studies at
Penn
Medicine and Mount Sinai Health System.
Synchron is
mapping the
brain via blood vessels. Inserted through the jugular vein, the Stentrode is
a neuroprosthesis placed in the superior sagittal sinus near the motor
cortex.
The eight-millimeter flexible alloy chip transmits neurological signals
to a
receiver unit implanted into the patient’s chest, which then translates
thoughts into clicks and keystrokes on a computer or mobile device in
real
time. Synchron is primed to start clinical trials and has teamed up with OpenAI to help paralyzed
patients
respond to text and chat messages.
Blackrock
Neurotech has
been testing its devices in humans since 2004 in its two decades of
brain-computer interface development. Blackrock’s product
portfolio has helped patients regain tactile function,
movement of
their own limbs and prosthetics as well as the ability to control
digital
devices solely from thought. Its latest project, Neuralace, is a flexible, hexagonal mesh patch
designed to
conform to the fissures and sulci of the brain. Its large surface area
can
capture 10,000 neural channels, inching closer to whole-brain data
capture.
Inbrain
Neuroelectronics aims to restore mobility in patients with
disabilities. The company has designed a graphene chip implant, which not only tracks brain
activity
but also stimulates it. In addition, it sends much more powerful
signals
compared to the metallic chips typically used in BCIs. While the graphene
chip will
be tested on a patient at the University of Manchester as part of a
brain tumor
surgery, it could later be applied to patients with Parkinson’s disease.
Following
a course
of neurological rehabilitation, brain chips and wearables can give
patients
direct control over exoskeletons and
robotic limbs. This is made possible
by reading signals directly from the brain, bypassing the site of
injury or
disease — such as a severed spinal cord — muscular activity
altogether.
With
brain-computer
interfaces, if you can think it, you can speak it. It’s just a matter
of how
fast neural decoding software can catch up.
A
team from
Stanford University found that its brain chip could hack 62 words
per minute,
which is on pace with natural conversation. The study featured a
non-verbal
patient who suffered amyotrophic lateral sclerosis and a pre-programmed
vocabulary of 125,000 words, marking “a feasible path forward for using
intracortical speech brain-computer interfaces to restore rapid
communication
to people with paralysis who can no longer speak.”
One
study noted that
individuals with ALS, cerebral palsy, brainstem stroke, spinal cord
injuries,
muscular dystrophies or chronic peripheral neuropathies may benefit
from BCIs.
Neural implants may be able to treat conditions or at least improve the
quality
of life for patients with chronic or terminal diagnoses.
Brain-computer
interfaces may one day be able to ease psychiatric conditions, like
bipolar
disorder, obsessive-compulsive disorder, depression and anxiety. Using neurofeedback
techniques, they can also help prevent more pedestrian conditions
like burnout
and fatigue by delivering targeted electrical stimulation to
specific
areas of the brain.
“It
may not look as
flashy, because someone in the demo just looks a little bit happier,”
said
Norman, whose research at the California Institute of Technology as a
postdoctoral
fellow focused on this next generation of brain-computer interface
tech. “But
if you were offered a solution where I said, using a single device, I
can treat
any form of anxiety that you have, and also offer you sleep on demand
and one
hundred other applications that could make your life a little better
than it
was, I do think that quite a few people would adopt that
technology.”
Users
can train
their brains — memory, executive function and processing speed — to the
biofeedback they receive from a neural implant in real time. Similar
to wearable tech and apps available today, users
would be
able to monitor their stats and self-regulate accordingly.
While
much of the
brain remains a mystery, BCIs are creating a direct
channel to
our thoughts — complete with a process to decode its language.
Brain-computer
interfaces have more than a half-century’s worth of research and
several proof
of concepts that passed human trials. So what’s the holdup? The two
largest
hurdles keeping BCIs from widespread adoption have to do with
regulatory
approval and funding.
Given
that
brain-computer interfaces are registered as a sort of medical
device, they lie under the jurisdiction of the FDA. Regardless of
the
product, the institution’s primary concern is patient safety.
The
challenge is in
the fact that BCIs currently exist in a league of their own. The
devices
themselves bring together a range of fields — implantable materials,
safety-critical software, the Internet of
Things and wearable medical devices, to
name a few — that are not yet standardized. There are no predicate
devices.
“These
are new
categories of devices,” Precision Neuroscience’s Mermel said. “So,
until
there’s a device that gets cleared for market by the FDA, it’s an open
question
of what type of evidence you have to show to demonstrate that the
benefit of a
device outweighs the risk.”
If
brain-computer interfaces
make it to medical practice, who is going to pay for them? Or for the
procedures? And what about the health-check follow-ups, ongoing
maintenance and
upgrades that support the technology over time?
Determining
whether
the tab gets picked up by healthcare and insurance companies,
government
subsidies or patients out of pocket will greatly determine the devices’
level
of accessibility to the public and who, based on socioeconomic status,
are
qualified for a tech-augmented quality of life.
“It’s
important to
prioritize the needs and perspectives of end-users, especially the most
vulnerable, such as those with disabilities, and consider the potential
ethical
implications of these technologies,” Alcaide said. “We must ensure that
these
technologies do not perpetuate social biases or further exacerbate
existing
inequalities