Page 20 - The 'X' Chronicles Newspaper - September 2023
P. 20
20 Brains May Predict The Future
To Make Sense of the
Present, Brains May
Predict The Future
Continued from Page 19
That is, the errors stemmed from what
was going on at higher levels of the
system — suggesting that lower levels
construct the error signal by comparing
incoming perceptions with predictions
descending from higher levels.
“It was exciting to find prediction errors,
and to find the specific content of
predictions, in that system,” said the
paper’s lead author, Caspar Schwiedrzik, majority of artificial intelligence research the University of Glasgow who has done
a neuroscientist at the European does not involve predictive coding, extensive work on the theory. But for
Neuroscience Institute Göttingen in instead focusing on other kinds of now, Sprevak noted, “the jury is still out.”
Germany. algorithms. []
According to Lucia Melloni, a researcher But formulating predictive coding PSYCHIC PREDICTIONS
at the Max Planck Institute for Empirical architecture in a deep-learning context
THAT NEVER CAME TO
Aesthetics in Frankfurt, Germany, her could bring machines closer to
PASS AND WHY
group is starting to see results compatible intelligence, Friston argues.
with an explanation of prediction error in
neuronal data currently being collected DeepMind’s GQN serves as a good • End of the world in 2012: This
from humans. example of that potential. And in 2017, prophecy gained widespread
researchers at the University of Sussex attention, but as we can see, the world
A Race to Find More Predictive even used virtual reality and artificial didn't come to an end as predicted.
Machines intelligence technologies that included
predictive coding features to create what • Presidential election outcomes:
Not everyone agrees that the case for they called the “Hallucination Machine,” Psychics have often tried to predict
predictive coding in the brain is a tool that was able to mimic the altered election results, and while some
strengthening. Some scientists accept that hallucinatory states typically caused by predictions have been correct, there
the theory can explain certain aspects of psychedelic drugs. have been many instances where their
forecasts proved inaccurate.
cognition but reject the idea that it could
explain everything. Others don’t concede Machine learning advances could be used
even that much. To David Heeger, a to provide new insights into what’s • Natural disasters: Psychic predictions
professor of psychology at New York happening in the brain by comparing how regarding large-scale natural disasters
University, it’s important to make a well predictive coding models perform like earthquakes or tsunamis have
distinction between “predictive coding,” against other techniques. At the very often missed the mark, as these events
which he says is about transmitting least, introducing predictive coding into are difficult to predict accurately.
information efficiently, and “predictive artificial systems could significantly
processing,” which he defines as improve the intelligence of those • Cure for diseases: Some psychics
prediction-making over time. “There’s a machines. have claimed to predict specific
breakthroughs in medical research,
lot of confusion in the literature because
these things have been assumed to all be But before that can happen, much work but these predictions haven't always
part of the same soup,” he said. “And lies ahead. Scientists need to continue the panned out.
that’s not necessarily the case, nor is it kind of research being done by Keller,
necessarily the best way to go forward in Schwiedrzik and others to determine just • Stock market predictions: Accurate
studying it.” Other types of Bayesian where, for instance, the brain’s internal predictions of stock market
models, for instance, might provide a representations reside. And it remains to fluctuations can be challenging, and
more accurate description of brain be seen whether similar experiments can while some psychics have claimed
function under certain circumstances. substantiate claims for predictive coding success, the field of finance is known
in higher cognitive processes. for its unpredictability.
What many experts in the field do agree
on, however, is that this research has the Predictive coding “is as important to Hummm: The question is “why do so
potential for exciting applications in neuroscience as evolution is to biology,” many people actually believe tat psuchics
machine learning. At present, the vast said Lars Muckli, a neurophysiologist at are real?” []