Michael Gillam, 6 November 2015
Quadriplegics have always faced immense employment challenges. They are plagued with co-morbid health conditions, can be socially isolated, and often live in near destitute conditions.
Today, a new breakthrough speed of 167 bits per minute was reported in non-invasive brain computer interfaces for humans typing words using their mind.
We decided that we wanted to answer two questions.
(1) Can we quantitatively predict the year that brain computer interfaces (BCI) will exceed the speed of human speech?
(2) Can we predict the year that BCI will enable the transmission of real-time video streams from our minds?
We discovered one answer and one new question.
How soon will humans type faster with their minds?
Human speech averages 3,000 bits per minute. Human typing averages 900 bits per minute. The world record for typing speed is about 5,000 bits per minute. In comparison, 320×240 video is roughly 250 kbps or 15 million bits per minute.
The question we were asking is how soon will technology bring us to 3,000 bits per minute via BCI – the rate of speed of average human speech.
We plotted the progress in bits per second across nine different studies from the top speeds across the last 25 years of non-invasive brain computer interfaces most of which used an approach called visual-evoked-potentials. We then fitted a trend-line.
The results were…disappointing (initially).
We were really hoping to see a realistic date when we might predict when videos of human imagination are projectable simply via thought.
Instead, the trend for these non-invasive BCI approaches revealed that we won’t reach the speech equivalence point even by the year 2030. We will have barely met the speed of regular typing.
Non-invasive brain computer interfaces using visual-evoked-potentials may never be realistic for typing at the speed of human speech before other technologies overtake them.
This led us to wonder if the sample size is large enough that the evidence might be strong enough to consider suggesting that these visual-evoked-potential based approaches to BCI may never succeed as a method to adequately capture and project human thought at equivalent bit rates to human speech given that other technologies have a chance to eclipse them.
We wondered if this type of trend line suggests we can look at a field and say with some confidence that a particular approach is non-viable and that new paradigms of non-invasive or invasive approaches should be tried. We have not ever heard that sort of advice being given based on quantitative exponential trends before – yet it seems like it is could be strategic advice valuable to researchers and companies. We imagine that this kind of pronouncement could anger or bother certain people for its seeming hubris. The core question seems to become – how confident can one be in the trend.
Then, we moved to plot invasive brain control interfaces where sensors are placed directly the cortex.
We found seven studies reporting bit rates. The data are preliminary but the trend could be promising.
If the current trends hold, the results could be transformative.
Given the current trends, BCI could equal human typing speed within the next 5 or so years.
Within 12-13 years, the average user of BCI will type faster with their mind than the current world record typing speed of 217 words per minute. By 2035, speaking via BCI can be expected to exceed every form of human speech.
Given current trends, within ~12 years the average quadriplegic will type 5x faster with their mind than we can with our hands.
If the trends holds true, then within 10-15 years, quadriplegics with brain control interfaces could become competitive in the job market.
Not only could quadriplegics be competitive, but they could become some of the most employable workers for certain professions because of how much faster they could write email with their brain computer interfaces. Quadriplegics could beat out their “average” human skilled coworkers for jobs.
The history of destitution that has often accompanied quadriplegia could become a relic of the past.
What an interesting time that could be.
—Mike Gillam, MD, FACEP 2015