Dean Radin MS PhD explains unconscious emotional precognition of negative events via Twitter hedonistics
Episode Summary:
Dr. Dean Radin, Chief Scientist at IONS, presents a radical speculation: collective sentiment might unconsciously sense and reflect future events. This hypothesis is grounded in the concept of precognition and presentiment, where individuals have a conscious or unconscious awareness of future happenings. Various experiments and meta-analyses support the existence of these phenomena, showing that people can have glimpses of future events under controlled conditions. Dr. Radin explores the potential of Twitter sentiment analysis as a tool for forecasting unpredictable future events. Twitter sentiment analysis has become a popular method for measuring collective emotions, used in various contexts, including advertising and market research. Dr. Radin collaborates with Peak Metrics on a project that analyzes words from diverse sources, including news, television, radio, and blogs, to understand and measure public sentiment. The analysis focuses on the emotional tone of the words, categorizing them as negative, neutral, or positive. Dr. Radin's speculation is based on the observation that collective sentiment might be influenced by upcoming events that are not consciously predictable, affecting people's moods and the language they use on social media platforms like Twitter. The document discusses the process of analyzing Twitter sentiment data, with a focus on predicting negative events. The analysis involves selecting tweets that reflect either very happy or very sad days, then creating an ensemble average of the sentiment around these days. The slope of the sentiment in the days leading up to the selected events is then analyzed to determine whether it can predict the occurrence of these events. The document suggests that there is a significant negative slope in collective sentiment approximately two weeks before unpredictable negative events occur. This observation implies that collective sentiment might be unconsciously sensing upcoming negative events. The analysis also considers the possibility of predicting not only when but also where and what kind of events might happen in the future. This prediction process is more complex, as it involves tagging each tweet with its geolocation data and analyzing the sentiment in different locations. Dr. Radin is currently working with a programmer to develop a method for pulling data directly from Twitter, along with geolocation information. The goal is to create a system that can predict when, where, and what kind of events might occur in the future based on Twitter sentiment analysis. The document concludes by highlighting the potential implications of this research. If successful, it could provide actionable insights into future events, allowing for better preparation and response to negative occurrences. The ability to predict the timing, location, and nature of future events based on collective sentiment analysis could have significant applications in various fields, including public safety, market research, and advertising.
Key Takeaways:
- Dr. Dean Radin from IONS presents a theory that collective sentiment on Twitter may unconsciously sense and reflect upcoming events.
- The speculation is grounded in the concepts of precognition and presentiment, suggesting individuals might have unconscious awareness of future events.
- Various experiments and meta-analyses support the existence of precognition and presentiment.
- Twitter sentiment analysis is explored as a tool for forecasting unpredictable future events, with a focus on negative events.
- The analysis involves measuring the slope of collective sentiment approximately two weeks before selected events.
- There is a significant negative slope in collective sentiment before unpredictable negative events, suggesting a possible unconscious sensing of these events.
- The document also explores the potential of predicting not only when but also where and what kind of events might happen, using geolocation data and refined text analysis from Twitter.
- If successful, this approach could provide actionable insights for better preparation and response to future events, with applications in various fields.
DEAN RADIN MS PHD EXPLAINS UNCONSCIOUS EMOTIONAL PRECOGNITION OF NEGATIVE EVENTS VIA TWITTER HEDONISTICS
So what's my radical speculation? It's that collective sentiment unconsciously feels the future.
And so while people may not be literally trying to predict anything, their mood might be affected by
something that is about to unfold in the future that is not predictable. And their mood would be
reflected then by the words that they use in describing anything that happens to be going on in their
lives.
- Dr. Dean Raiden MD. PhD. - Chief Scientist of IONS
And welcome to Connections Live, which is our free weekly webinar series which is designed to help
you deepen your sense of interconnection, to inform and inspire. I'm just incredibly delighted to
introduce Dr. Dean Raiden. I'm sure, sure that many of you are familiar with Dean's work or have read
a book or seen one of his presentations before. But for those of you who are not familiar with Dean, or
maybe this is your first event here with us at Ions, I'd love to just read out a short bio.
So dean is Archie scientist here at Institute of Knowledge Sciences and associated Distinguished
Professor at the California Institute of Integral Studies. Before joining the research staff at Ions in 2001,
he held appointments at at Amp, T, Bell Labs, princeton University, university of Edinburgh, and Sri
International. Dr. Raydon is co author of hundreds of technical articles, more than 125 peerreviewed
journal articles, four dozen book chapters, and four bestselling popular books, including The Conscious
Universe, Entangled, Mind, Super Normal, and his most recent book, Real Magic. And he has given
over 600 invited presentations and interviews for government, military, business, scientific and other
groups all the way around the world.
And I'm guessing this number is much more than 600 because it's been on the bio for a while. So,
Dean, so appreciate you being here with us today and actually sharing some new research that you are
up to around how we can sense into the collective field of consciousness and maybe see something
about recognition. So from there, I will let you take it away and everyone can be thinking about your
questions for Dean as well. And at any point put them in the Q and A box, we'll circle back to them. So,
Dean, welcome.
It's great to have you here. Thank you. Thank you, Andrea. Okay, here we go.
So this is about Twitter sentiment and pre-sentiment.
So to forecast where we're going, since we're talking about precognition, the question is, can we use
tweets to forecast future events that are unpredictable? What is sentiment analysis? It's analysis of
words or text to measure negative, neutral and positive emotions. We can do that using tweets. We have
a happy bird and sad bird.
Twitter sentiment analysis is extremely popular now. It is being used for advertising agencies that want
to know whether a promotion that they're doing is reaching people and how people respond to it, but
many, many other contexts as well. We're working on another project with a company called Peak
Metrics, and Twitter is one of the sources of the words that they're using, but lots and lots of other
sources, everything from the news to television, radio, blogs and so on. So any place that produces
something that is either in words or it can be turned into words through a transcript can be used as a
reflection of what people are thinking about. But more importantly, the sentiment measure is really
looking at what emotions are about happy and sad emotions.
Why do I think that Tweets might be able to forecast the future? Well, because we know that
precognition exists from an empirical perspective. In precognition, the word cognition is implying that
it's about preknowing. It's about having conscious awareness of a future event. Well, how do we know
that?
Well, there's a bunch of experiments. There are forced choice experiments, there are implicit
precognition experiments, there's precognition and dreams. There are lots of different kinds of evidence
showing that under controlled conditions, people can get a glimpse, usually a tiny glimpse, but a
glimpse about future events. But someone could then say, well, but Twitter users aren't trying to predict
the future, which is true, but we also know that presentiment exists pre sentiment, meaning prefeeling,
which is an unconscious effect.
How do we know that? Because of lots of experiments using physiological methods to look at what's
happening in the unconscious. And the metaanalyses are very clear on these studies. Lots of meta
analyses have already been done showing that part of our unconscious is reflected in our physiology
does feel the future, and generally it feels negative futures, more than positive or neutral futures. So
there's there's something there in each of us.
Well, the world is a strange place, and nothing but radical speculation gives us the hope of coming up
with any candidates for the truth. This was said by Thomas Nagal. The philosopher has written a lot
about the philosophy and biological studies and living systems. So what's my radical speculation? It's
that collective sentiment unconsciously feels the future.
And so while people may not be literally trying to predict anything, their mood might be affected by
something that is about to unfold in the future that is not predictable. And their mood would be
reflected then by the words that they use in describing anything that happens to be going on in their
lives. This is a work in progress. I'm still in the midst of looking at this and in early stages. So this is
not published yet.
This is work I'm actively working on now. So the source of the Twitter sentiment data comes from the
University of Vermont. Within their complex systems center, they have a project called Computational
Story Lab. And so one aspect of this project has a site you can look at, which is
Hedonometer. Hedonometer.org, of course, hedonism is all about the pursuit of happiness.
So he denominator is a coined term for measuring happiness, and that's what this site does. If you go to
this site, you'll see a page that looks something like this is showing the average happiness each day
based on Tweets, in this case in English, but they have ten different languages. So what the graph is
showing is the measure of happiness from one day to the next. They've annotated some of the dates.
They also color code the day of the week, and among other things, you can see here that some days will
generally be higher than others.
So the weekends tend to be happier than Monday and Tuesday, for example. And you can see here, like
Mondays tend to be these purple dots here, which are usually a little bit lower, and the red ones are
Saturday, which tend to be higher. So there's a lot of predictable effects that happen in these tweets.
What is not predictable is when something negative happens. So in the bottom here is the volume of the
tweets.
So at this point, the volume is somewhere in the vicinity of 200 million tweets per day, just in English.
And if you look worldwide, it's a lot more than that. So there's a lot of data available.
So these are the ten languages that they have. They have a bunch of other projects too, but I'm focusing
mainly on the Twitter languages. That sentiment has been analyzed already in ten different languages.
So those are the languages, this is how they analyze the sentiment. So you have a number from one to
ten, which is degree of happiness.
So these all happen to be English words that are happy, and they all get pretty high happiness scores,
but they have 100 different words. And so people have gone through these and then they take the
average of what people think is how happy or sad is this particular word. So they already have that built
in. That's how they figure out based on the language and the tweets, is this collectively, is this a happy
day or a sad day? It's a little bit more complicated than that, but they go through in pretty good detail
on this website, in their papers and their descriptions on how they come up with the sentiment metric.
So here's from 2009, which is when it began to today, you can see that there's some very happy days or
some unpredictable negative days. The positive sentiment days are almost always planned events. So
Mother's Day and New Year's Eve and Valentine's Day and Christmas are these giant spikes over and
over again. The negative sentiment are almost always unexpected events. Things like earthquakes,
terrorist bombing, death of a celebrity, a mass shooting invasion of Ukraine by Russia, which was like
people didn't know they were actually do that invasion until it happened.
So the positive sentiment, planned events. It would not be surprising at all if there was some forecast to
it, because people are anticipating that something pleasant is going to happen. But you would not
expect that there should be a trend before these negative events. And there's again, the number of daily
tweets. So Twitter started out small.
It is now big and uncertain. Events like these were protests against police brutality. You get a huge
spike in the number of people who are doing tweets. So here's the original English tweet over almost 50
days. And so I create a moving mean to smooth it out, 30 day moving mean.
And you subtract the two and you end up with a pretty nicely behaved sequence here, which you can
also then normalize and put it into z space and you can pull out the most happy days and the most sad
days using a threshold. So that's how we pull out our positive and negative events that we're going to
try to predict. Also, when something negative happens and also when things positive happen, there
tends to be a delay. Some people will continue to talk about these events. So I don't select any days that
are within three days of each other for these selected events just to get around this idea that something
bad will happen and there's a lot of tweets about it, but by around four days later it goes back to normal
again.
So here's an example using Portuguese tweets because it's a pretty clear example. So what we do here is
this is the day where it's selected as being particularly happy and this is the day is particularly sad. We
take those days and then 14 days on either side of that and we create an ensemble average. So this is the
average happiness around for two weeks on either side for very happy days and very sad days. The
question then is when we do a slope from 14 days up to twelve days before, does that actually begin to
predict the sad event?
And the reason why we don't go one day before is because of time zones and when the day switches.
So you want to be sure that people are not already that some people in some time zones don't already
know the negative event. So we go two days before and that prevents the data from being
contaminated. As you can see here already that just before the sad day, there's already a drop
happening. So that means some people in some time zones are already in that second day.
So again, we're looking at 14 days around selected tweets. Those are the happy events, those are the sad
events. And the question then is are these slopes significant? So again, the positive slope would not be
surprising because people know that the positive thing is going to happen.
This peak here, which you see in some languages, not all languages are typically that one week. This is
seven days before the positive event. People are anticipating christmas is happening next week and
Valentine's Day is coming up and so on. So that's why you would see this peak for sad days. There
would not be a peak because people don't have any reason to have a peak because they don't know that
the event is coming up.
So how do we figure out whether that slope is significant? Well, you can calculate the slope for this.
Usually they can be quite significant, but that’s because there is a dependency among each of these
dots. So you need to take a slightly more sophisticated statistical approach in order to see whether that
slope is actually significant or not. So we use a permutation method.
And the way it works is, first of all, we select the most happy and the most sad days. We create a slope,
find a probability of that slope for each one of these days and focusing mostly on the sad days. And
then you take all of the original Tweet data and you circular shift it by a random amount. And what this
is doing that is saying we’ll keep all of the original dependencies in the data, but we’re just going to
shift it so that as though the Tweets occurred on different days, we calculate again 14 days plus or
minus the original day, calculating the new slope, got a new random score or a Zscore. Repeat this a
thousand times and then the final statistics based on the Zscore is our original minus the mean divided
by standard deviation.
So this is a pretty standard way of figuring out whether an effect in time series data that has
dependencies in the data is the way of handling those dependencies in a nonparametric way. So the
zscores we end up with are valid. So for English Tweets, the slope associated with this line that looks
almost completely flat and the Zscore is negative, but it’s not significant. For French you can see more
clearly it’s a negative slope. And even after the permutation test, it is still a significantly negative slope
up to two days before the negative events.
Spanish is also significantly negative, ukrainian is negative, not quite significant, korean is somewhat
positive, german is significant, negative, arabic is negative, not quite significant, russian is negative,
indonesian is significantly negative and Portuguese is significantly negative. So when you add all of it
together, you have the Zscore for the sad events. This is two weeks before the unpredictable negative
event, highly significant negative slopes. This is suggesting that about two weeks before people are
feeling something. You can do the same thing for happy events.
Again, as you would expect this to be positive and in fact it is probabilistically it's certain. So two
weeks before an unpredictable negative event, collective sentiment looks like it is feeling the future.
Well, how far in advance can we predict this? So you can do these predictions where you're starting
with say, just five days in advance and always leaving two days. So you go out two days and then look
for a slope of five days long.
So I did that for five days up to 30 days. It just so happens that 14 days is the optimum. That’s where
you get the largest effect. So that’s interesting. I’ve done a bunch of other analyses too, but they’re not
quite as far along as what I’ve done so far.
So I’m not going to talk too much about all of that. So the implications here are that it looks like
presentiment on an individual basis scales up and is reflected in daily collective sentiment of millions
of people that’s one way of thinking about it. I think that is what’s going on. And so retrospectively,
which is the analysis that I’ve done so far, we have all the data, we’re going back and looking at it
retrospectively, the sentiment seems to anticipate when something sad is going to happen. And of
course, again, to emphasize here that these are events that cannot be predicted or inferred.
We know that bad things are going to happen, but we don’t know when they’re going to happen. So
there’s no reason for Twitter to reflect that unless people are feeling that something sad is going to
happen in the future. The more important question is, can we predict when something sad is going to
happen? In other words, not retrospectively, but prospectively. So retrospectively is interesting from a
scientific perspective.
Prospectively becomes pragmatically useful, and that’s where I’m moving. Then in this kind of
analysis, the prediction would allow us to say that as new data comes in every day from Twitter, if we
can keep track of the direction of the slope, of the sentiment values, that when the slope reaches a
certain degree of negativity, we can then predict that two weeks later something bad is going to happen.
Or maybe less than two weeks, but at least before it happens. The second thing is, can we predict where
the event is going to happen? It is interesting and maybe pragmatically useful when something is going
to occur, but for anybody to actually do anything about it, they also need to know where.
And so fortunately, Twitter data also has geolocation data. We know where the person was when they
sent their tweet. It’s a much more complicated analytical problem because now we need to tag each
tweet with its location. And so when you create a sentiment data, you’re going to have sentiment that’s
different scattered all over the place. So it’s a much more complicated problem.
I’m now in the process of working with a programmer to see if we can create a method where we’re not
using the data from the Heedometer.org site. We’re pulling the data directly from Twitter along with the
geolocation data, and then figure out if we can predict both when and where, and even better, would be
able to predict what? So what would require a more refined form of text analysis, because now it’s not
simply emotion, but it’s to see whether or not certain kinds of events have different kind of words that
are associated with it. So a couple of years ago there was this mass shooting in Las Vegas where a lot of
people were shot. And at the time, that was the most negative event on Twitter’s history.
And what I was interested in, that it was looking at a week’s worth of Twitter before that event and a
week afterwards, that the nature of the words begin to change. Not the sentiment so much, but the kinds
of words that people were talking about. And there saw indications that the words had begun to change
in the direction of something like shooting. Well, again, if we can figure out when, where, and what,
then we have something which is actionable, and that makes it much more interesting. So that’s what
I’ll be moving towards.
So you might remember that the Philip K. Dick story, Minority Report, which was made into this
movie, it was all predicated on this idea. We can predict what, when, and where, and intervene and
maybe stop a negative event from unfolding. Of course, that could create a paradox. And I think the
way they got around this paradox in The Minority Report is that they could stop a major event, but only
just before it happened.
In other words, just that somebody was planning to do something that wasn’t quite enough in the
movie. That’s how they made the story. But I think in the real world, we might be able to intervene, but
not stop it to the extent where it wouldn’t have shown up in the first place, because that would be a
paradox. This is all about the grandfather paradox and time travel. So we’re talking about creating a
department of preprime for real.
Well, ten years ago actually, it was twelve years ago already, the DARPA, the Defense Advanced
Research Projects Agency in the US. Had a program called Anomaly Detection at Multiple Scales, and
this was all about using social media data and sentiment values to see if they could predict, if anybody
could predict. This was a call for proposals for people doing this kind of analysis. And there was $30
million that applied to that, and people did things, and nothing came of it, or something did come of it
come up, and they classified it. So we don’t know what happened, but it seems to have disappeared.
It’s not on the radar anymore. I would be a little bit surprised if nobody came up with something
interesting, because as far as I could tell, I was able to come up with what I’ve done so far fairly
quickly. The difficulty ten years ago was that sentiment analysis wasn’t quite as good as it is today, so
maybe they didn’t have the tools. Now you can do it automatically, completely automatically by the
computer, and it’s becoming very, very useful for a lot of applications. At the same time, it was known,
because this is a public request for proposals from DARPA, that maybe you could use Twitter to predict
the future.
What almost all of that meant, and continues to mean in the forecasting business is if you ask people,
what do you think of this? And you take advantage of the crowd guessing what they think is going to
happen, can you use it to predict stocks, and can you use to predict events and so on. There are still
companies that are doing that. It sort of works, but it’s very different to what I’m talking about, because
I’m saying that our mood, our collective mood is feeling something that is unpredictable in the future
and unconscious, which is an important element here. We’re not asking people to predict anything.
We’re simply looking at their behavior and using that as a way to predict what is about to unfold.
So I’m going to stop there and I thank you for your kind attention. I think we’ll have a lot of time for
questions.
Great. Yes. Thank you, Dean, and invite you all to post your questions in the Q and A box for Dean and
we’ll get started. And I want to start out with one question for you, Dean, about what’s kind of the
theory, you know, normal materialist science says that’s not possible to predict something in the future
or that there’s something like a collective consciousness. I know there’s different theories, but kind of
what’s your leading edge theory of how this works within postmaterials scientific worldview?
Well, it depends on what the phenomenon is that we’re talking about. And some of the research we do,
like the Global Consciousness Project, we’re looking at a mind matter interaction. And in that case,
anytime we’re talking about a direct connection between mind and matter. If you look at that from the
point of view of reductive materialism, which is the main theory in science today, it’s very difficult to
understand how that could happen. So we need to look at other worldviews, other philosophical ways
of thinking about such things, and the two that are most easily accommodate the notion that mind and
matter can interact.
One would be idealism, which says that consciousness is fundamental. It’s more fundamental in the
physical world. The other approach is dual aspect monism, which says that the physical world is quite
real, it’s out there, it’s part of the fabric of reality, but the mental world is also quite real. And so these
two aspects of reality, then the matter side and the mind side, are imagined as emerging out of
something even more fundamental, which Carl Jung called the UNUS mundas, the one world, the one
thing which is completely holistic, out of which emerges mind and matter. So in today’s world, we’re
thinking of, well, what could that undist menundas be?
Well, we can’t know it directly because we live in a world with mind and matter. Some scientists and
scholars now are suggesting that it is a world of information. It’s like pure information. It’s not quite
physical, it’s not quite mental, it’s somehow mixed or both. So there’s an informational reality out there
from which the world, as we experience it, seems to arise.
The important thing about that particular world view is that mind and matter then become two sides of
the same coin, and as such, they become intimately related to each other. So if something happens in
the mental space, there has to be a reflection in the physical space. And in the neurosciences, that’s
exactly what you see. We’re seeing neural correlates of consciousness the brain is doing things as the
subjective mind is doing things. But this would say that it doesn’t stop with the brain and the mind.
It’s everywhere because now the mind part, if a lot of minds suddenly become very coherent, then
some aspect of the physical world is going to become coherent as well. So that’s where you need to
think about alternative worldviews. For what I’m talking about here, all we need is the possibility that
we can feel future events so we could still stick with materialism provided that we can entertain the
idea of retrocausality. So retro causality is okay in most physics in terms of the equations of physics are
times symmetric in most cases. So it’s not inconceivable that there is a retro causal element in our
experience and be able to feel the future as we see even in laboratory experiments.
But this is now on a very large scale and we’re talking then about changes in mood. And of course, as I
said, the real important question here is can we use this in a pragmatic way? So that was my original
impetus to look at these data, to see can we figure out how to look at these data in a way so we can tell
who, what, where, when? All of the rest I’m not sure we can do who because of who in this case is all
of us. But the other questions we might be able to answer absolutely.
That’s fascinating and I love the different explanations and even in materialist, more physics equations
that something like this could fit into the paradigm as we’re looking at it. That’s fascinating. And
actually a really interesting question here from Peter. He says we might guess that it is the negative
event that causes the reduced presentiment mood. But has anyone looked at whether a decline in mood
can cause the negative event?
So kind of looking at that coherence or not in the field causing something in the world or vice versa. So
what would you say to that? Yeah, this is a question that comes up a lot in thinking about the
precognition versus psychokinetic events. Are we causing the event or how do we do that? How do we
even think about it?
Well, for some kind of negative events like the death of a celebrity, so we think that the movement of
mood pushed by whatever reason, I mean it could be pushed by a lot of things. But does the collective
negative mood result in the death of a celebrity? Well, possibly, I guess. I hope not because that would
mean that any time the large scale mood is beginning to change that all kinds of negative things are
going to happen. And maybe we only see the ones that are like a death of a celebrity because it happens
to go into the media.
But could it affect everyone? Well, yeah, I guess so, right? I mean, there are other studies looking at the
mind matter interaction and suggesting that collective attention alone will create physical effects. So
there may be other things going on. So we know that solar wind is a correlate of some of these
phenomena and the geomagnetic field and maybe other effects that modulate these things.
There’s even evidence that sunspots modulate violence and more. So we’re being pushed around by all
kinds of complicated variables here. It’s usually not possible to know why something is unfolding and
in this case, from a scientific perspective, it would be very interesting to figure all that out. But again,
my push on this is pragmatic. Can we actually tell something in advance in a way that can be useful
and maybe at the same time we can be smart enough to figure out why it’s all working but I don’t think
we actually are smart enough to figure that out yet, we just know it’s an interesting question, right?
Seeing some symbols and signs of something happening in the larger field yeah, fascinating. And
Shelley’s asking some questions more about the individual. So talking about the collective field and she
says I’m curious about how it’s possible to see inside someone’s home never even met some people
she’s never even met, she calls them flashes, seeing people doing things and then within 1820 to 24
hours sees that person in the news or earthquake happen. So it seems just having an experience of
seeing something before it happens, how common is that? Or in your experience kind of this
precognition research, maybe you can speak to that a little bit?
Well, it’s common enough so that when we do experiments in the laboratory we can even under well
controlled conditions, we can see that people have disability. The next question most people want to
know then well yeah, it’s a real phenomenon. One of the most popular kind of spontaneous psychic
effects that occur are precognitions and dreams. So it happens, it’s there, but then people went, oh well,
how is it there? How is this possible?
Well, we don’t know yet exactly. We have to use terms like retro causation and there’s some physical
reasons to believe that but we know so little at this point about how does that information get into us? It
requires almost that part of our awareness is spread out in time and again, from a pragmatic
perspective, it doesn’t actually matter yet for us to understand theoretically why this is working, it’s
more useful at this point. And the reason why I’m focusing on the pragmatics is that the amount of
research funding available to do these kinds of studies is very low, like really low. In which case in
order to get people excited about it, you have to show not simply that it exists, which is scientifically
interesting, but that it can do something and then people, when there is moment that they can see that
this is actually useful, well, then funding would flow.
We’re moving in that direction but. That’s why then this particular project is looking at this particular
social media as a way of making these predictions, right? Yeah. So how is it usable in our world, either
for ourselves or into the world? And Russell is asking, might preventing a precognized event prevent
pre sentiment?
So it’s kind of that time effect. What do you say to that? Yeah, that’s the paradox I was referring to. If
you know that something is going to happen and you stop it, how did you know that in the first place?
Well, one of the ways of getting around that is that we imagine that the future is not fixed.
The reason that you’re getting information from the future is that it does not faded to occur. It’s not
deterministic, but rather probabilistic. So at any given time there is a probability that some event is
going to happen. And maybe what we’re picking up then are the probabilities of it. So even in the case
of a collective presentiment, maybe there’s a collective change in mood which is reflecting a
probability that something is about to occur.
So if it turned out that you had negative events happening like every two weeks, like a clock time, well,
then it’s no longer unpredictable, it would be predicted, but the probabilities of it would also be
extremely high because anytime you have a predicted event, obviously it’s going to happen. So it’s not
surprising. But that’s why the events that I’m looking at here are ones that would not be predictable. So
I don’t know, is this a case of people sensing the probability of something happening or feeling of an
event that actually occurred? There are a number of experiments have looked at this issue of is the
future fixed as invaded or is it probabilistic?
So I’ve done experiments in this, a few colleagues have done this. All of the experiments that I’ve done
have suggested that the future is probabilistic, something will probably occur. And then that gets
around the paradox too, because you could have something that occurs which is not quite as bad as it
would have been if you hadn’t known about it in advance. So this is relevant then to the notion of pre
crime. We wanted to stop a crime.
You need to have something happen in the future in order for us to have the prediction, but maybe we
can prevent it from being as bad as it would have been otherwise. So that’s the way I’m currently
thinking about it, because otherwise it makes your brain hurt when you start thinking about the paradox
that comes about and trying to stop something completely, in which case nothing would have come
back in the first place. Right? Yeah. Bringing to mind all kinds of different scifi movies and plays on
that time loop.
So it’s a really great question. And there’s a couple of other specific ones about this specific study
about why only Twitter, why not other things like Facebook did you have a reason why you chose
Twitter? Yeah, because Twitter has an API that allows you to draw the data. It wasn’t always free, but
now you can get the data for free. So it’s just a matter of, I don’t have any funding for this.
This is not a funded project, let’s put it that way. So maybe that’s why makes sense. And Ek is asking,
did you notice the same pattern with no event following? So maybe where sentiment was dropping, but
there was no negative event? Have you looked at something like that?
Yeah, that does happen. You do see long term fluctuations in mood and you see short term fluctuations,
and that’s the reason why it’s pulling out only the most negative events. So that’s why I had this
thresholding method where pulling out the most happy and the most sad, because I figured that if
people were really picking up something about a future negative event, you want to get the extreme
contrast there in order to be able to see it. Because even like we see in the laboratory, the effects that we
see are pretty small in magnitude, and so you want to optimize the chances of actually picking
something up. And the other thing is that these negative events, and the positive events too, they tend to
be carried by the media.
So when the event occurs, a lot of people are paying attention to that event. That’s reflected in Twitter
and Facebook and everywhere else. But I think there’s something about a lot of people paying attention
to that event that is perhaps something like the source, the reason why we’re all feeling it. Most of the
time, Twitter is sort of fluctuating around zero because there are people talking about what’s going on
in their daily lives. Occasionally something pops up in the media that attracts a lot of attention.
That’s the thing which I think is the actual source of this information. It’s not this ongoing noise,
essentially, which is just individual human events. Yeah, absolutely. I’m seeing a follow up from
Shelley also asking about if it’s stressful, if certain precognition things are coming and it’s stressful to
the individual. Have you seen that?
Or do you know places of how to work with this information as people are experiencing it? Well, the
usual concern here is that somebody saw something and didn’t intervene because maybe they didn’t
know how or who to contact and it happened and then they feel guilty about it, like an airplane crash or
something like that. If they figure, if I only had told somebody. Well, as you can imagine, people in law
enforcement and elsewhere are getting tips all the time and there’s simply not enough people around to
follow up everything. So I would say that at this point, I would say the evidence is not that strong,
suggesting that a precognitive event means that you’re causing that event.
I think it’s more like we’re perceiving something is happening, but not causing it. So since we can’t do
very much about that at this point, I would say, well, don’t worry about it because if you get the
information, you could send it up somewhere. There is a slight danger of doing that. If you make a
prediction of some kind of event that is a negative event and the event happens, you may be blamed for
that event. So this is particularly sensitive when it comes to things like terrorist activity because from a
conventional perspective they’re saying, how in the world could you know that?
The only way you can know that is if you were part of the reason for why the thing unfolded in the first
place. So for years there was, based in the UK, a central premonitions registry, which is a way to
anonymously give tips about precognitions that are occurring in a centralized place, which would be
useful because if a lot of people had similar predictions, you would have some reason than to go to law
enforcement and say, well, a lot of people are feeling something bad is going to happen at this
particular thing. That would be a cleaner and safer way of dealing with these kinds of information.
Because it’s true that I see from the questions that are already posted that people do get anticipations
about things that in fact do occur. But what do you do with it?
Well, we were thinking about creating a website that would have a new form of web based central
premonitions and then we started thinking about the liability associated with it. Even if it’s anonymous,
we’re starting to think that if we had a bunch of data suggesting that a certain plane crash was going to
happen, are we liable for that? Even the people who are collecting the data suddenly now have a certain
responsibility for these future events. And we decided after thinking about it for a while, that we didn’t
want to take on that responsibility. So I know there’s some people putting in Q and A here who have
had very interesting precognitions and have intervened by talking to people in law enforcement and
sometimes they do actually get effects that are useful.
It’s tricky though, like what do you do with all that? Yeah, see, as Ben is saying, yeah, you end up
crying in the shower and tell yourself you just get up and go again. That’s how you work it.
When we start taking it seriously, looking at it from a scientific and psychological and even around
governance and how we keep everyone safe, it’s really fascinating when we are able to open up all the
implications to this if we can sense into the future in the collective field. Yeah. So David is asking is
this experiment that you’re doing basically like the web bot experiment in the 2000s, looking at words
on the internet to predict the future? I don’t know if you know about that experiment and now is Twitter
the better way to do that? Yeah, so I do know about that, and it was quite interesting.
What I’m doing, which is quite different than that, is just looking at the sentiment of the words, right?
Not using the words as a way of predicting somebody’s going to say something, and then that word is
meaningful in the future. That would be necessary at the stage where you’re trying to figure out what is
the future event. So that will be part of this eventually. It’s much more difficult now than it used to be,
because when you look at something like a Twitter feed, you’re getting 10% of a randomly selected,
10% of the actual full data set.
But that’s already gigantic. It’s huge. So you have to build a fully automated way of doing this, which
means that you you don’t want to have the luxury, in a sense, of looking at individual words and trying
to see a pattern in it. It has to be automated. So a combination of algorithms and AI and other methods
are necessary in order to be able to do this.
That’s why in the hedometer site, that is completely automated, you can tell because the day after well,
it is up to date as of yesterday. So they’re pulling in all the data for one day. They crunch the data, and
then the chart is updated every day. So something like that would be necessary for this, where most of
the time, fortunately, there aren’t events that we care about so much. But eventually something is going
to happen.
And the algorithm has been looking at how the data is going and saying, well, there’s a lot of negativity
showing up in these tweets which are not based on anything as far as we can tell. So is Twitter the best
way? Well, as we already said, the data is free. You can get data from other sources. I think Reddit has
an API.
Few other places have APIs for a funded project. I would look at many, many different sources, like the
company peak metrics. They’re looking at a huge range, including even things like YouTube and Tik
Tok. So you would translate the words, translate the videos into transcripts, and you have those words,
and you can do the same thing with that. It would actually be really, really good to have multiple
sources.
So they have a way of testing that. This effect is not just seen in Twitter, but it’s a generalized effect
across social media. That would be great. That’s ideally what you would do. It would cost a lot to do
that, right?
Yeah. Sharon Joy is asking, how might such experiments connect with entanglement? With Nobel Prize
in physics given to quantum scientists might just become more attractive to funders. So how do these
things relate entanglement and precarion? Well, I don’t know that this would be directly relevant for
entanglement other than it appears from these kinds of experiments that minds are entangled.
There’s two ways of thinking about it. One is there’s a lot of people, millions or even billions of people
thinking about and anticipating something at the same time. So we’re picking up a scaled up effect. The
other way is that there’s just one mind, and we’re all little pieces of it. We’re little holographic pieces of
this giant mind thing.
It doesn’t really matter which one of those is the correct answer. What does matter is to not push this
too far is the pragmatics of it at this point. And this is not that unusual within science that people get
really excited when you can show that something is real. So entanglement, when it first was being
studied, as the reports have said, when Klausar was doing these experiments, he was warned not to do
them because it was considered to be mystical nonsense that physics didn’t care about. Hardly anybody
was looking at the fundamentals of quantum mechanics at the time.
So at the beginning of new knowledge, it’s very difficult for multiple reasons a lack of funding, a lack
of people’s interests, all kinds of things. So you kind of need to show why anybody would care about
this. Well, now, entanglement is extremely important because it’s being used in a variety of
technologies and is very important in the development of quantum computing. So huge amounts of
funding are flowing. But as I said back in the 70s, when Plaza was first doing this, there was zero
funding, and they were doing dumpster diving in order to get the equipment to do the experiment.
So I’m kind of at that same stage, I’m doing dumpster diving into Twitter feeds to see whether or not
we can come up with something that would excite people. Yeah, excite people. And like you said, to
make it pragmatic to be used. So that also funding and more interests can come into the field too, right?
Yeah.
Nate is asking a follow up question on the entanglement. One is, can’t mass meditations or healing
events account for a positive mood? Is there some research about that? Well, that’s a whole other talk
that’s about the global consciousness project and also studies that have been done by the transcendental
meditation organization looking at the effects of largescale meditations, not necessarily meditations for
peace, but just meditations, calming groups. So, yeah, there’s actually a fair amount of data suggesting
that the combination of meditation, coherent attention radiates out in some way from that group and
affects people in the vicinity by reducing violence and changing their mood.
But again, that’s a little bit more on the side of a mind matter interaction or perhaps that we’re mentally
connected in a way so that if there’s an enormous amount of agitation somewhere, we will feel more
agitated and vice versa. If there’s a lot of calm going on, we will feel calmer. So it’s not that unusual for
people who do meditation. If you do meditation by yourself, you can experience it a certain way. If you
meditate in a group, especially of experienced meditators, it’s a very dramatically different feeling.
So there’s there’s some sort of field effect that seems to go on in these kinds of collective experiences.
And that’s why I think something like that is going on. Even in predictions, we can feel the future
happening, where will happen? I don’t know what tends to use. Yes.
And do you think if we were another mass meditation, something like that would show up in although
it would be predicted event, do you think it would show up in the Twitter data or have you began
looking for planned positive events outside of Christmas and things like that? Well, anything that is
planned then the idea of anticipating it doesn’t make any sense. Right, because if it’s planned but it’s
secretly planned well, that would be interesting. That would be a test yeah. To see whether mood is
beginning to rise.
That kind of data would be very difficult to pick up on Twitter unless there was a group somewhere that
saying we’re going to get it, secretly get a thousand meditators together and do something. But nobody
knows about this other than those 1000 people. Well, yes, that would be a very interesting experiment
to do, because you could predict based on this kind of experiment, that Mood sentiment would begin to
actually rise towards that event. Now, the TM people say that you need a certain number of people in
order for this effect to occur, like the square root of 1% of the population or something like that. So
maybe 1000 is not enough.
But this again is where geolocation data would become very useful because maybe 1000 people would
affect the local area, but not the world at large. And so if we knew that, we could see what was
happening to people who are tweeting, who did not know about this big Meditation for Peace event.
Were there tweets affected? Well, we can in principle figure that out, right? Absolutely.
And it’s great that we can get the geolocation and see is there a local effect or not for some of these
events. Yeah. And Joe is asking, are there any particularly promising use cases for this phenomena as
an industrial application? You spoke a little to it, but you’re saying like predicting stock prices or
something else. Are there some good experiments or use cases out there right now?
Well, what people are using sentiment for now is prediction of the stock market and financial
instruments, as I said, success or failure of advertising campaigns, politics, any area where you’d want
to know is something working, especially if it’s working in the future. There are ways that people are
using it sentiment data for that, to my knowledge. I don’t know anyone else who’s been using
especially Twitter sentiment for making the kind of predictions that I’m interested in, because I’m
looking two weeks in advance and that seems to be the optimal amount. And I do have preliminary data
that I’ve been working on in the last couple of weeks to see if we can predict, literally prospectively
predict rather than retrospectively. And the answer seems to be yes.
So when the slope becomes particularly steep, it does seem to predict. And of course this is very
important because retrospectively is interesting, but if the prospective one would mean that we just
need to keep track of what the slopes are from day to day. And also when we know there’s a planned
event coming up that if we start seeing a positive slope, we’ll say, well, that’s because it’s going to be
Christmas in a week. If there’s a negative event and it’s not around some day of remembrance or
something that’s going to be sad. There is evidence that I’m seeing that, yes, there is a higher than
chance effect that what is going to happen is going to be very negative in somewhere between two and
14 days hence.
So then the next step, as I was saying, was where is that going to be and then when or what? Those are
the two remaining questions and I’m just now beginning to work on those. Wow. Yeah, that’s
fascinating. It’s really fascinating to see how we can use this real time information from Twitter,
massive amounts of information, to help give us a clue to what’s happening in the world.
It’s incredibly fascinating and I do have one more question for you, but I do want to give a couple of
announcements before and then we’ll wrap back to Dean’s final thoughts for the day. So I just wanted
to let you all know that next week we’ll be taking a break from our free webinars but for launching a
new special membersonly event called Interconnection. So if you are an Ions member, check your
email. And this is going to be an event designed to really foster community and actually be able to meet
and share with each other in small breakout groups. And we really heard from you that you wanted
more engagement and community.
So we are very excited to be launching that event next Friday. If you’re not currently a member, we’ll
send out that link. We’d love to have you join and we’ll be doing more and more of these kind of more
intimate community events for our members. And if you are a member, check your email for the link to
register. So we hope that you will join us in and then the following week we’ll be back live with our
connections.
Live at 11:00, a.m. Pacific for a panel discussing Earthbound Spirits and around the world. There are
cultures that believe that a person’s spirit can be trapped on Earth after they die and that phenomenon is
referred to as Earthbound Spirits. So if you’re curious to hear more, we have a great panel lined up
where our guests will share insights on from a scientific, clinical and personal perspective about this
phenomenon. So definitely join it’s right around Halloween too.
So we get to really sense into what is the current science and psychological perspectives about some of
these concepts that have been with us for a very long time in many cultures across the world. So we
hope you will join us there. And before we turn back to Dean for his final thoughts, I’d just like to
mention that these free programs as well as our scientific research are made possible with the generous
gifts from our members and donors. So just a huge thank you to everyone on the call who participates,
who chats, who donates, any way that you’re participating in this work. We are so grateful and we
couldn’t do it without you.
And if you are inspired, we invite you to go to Noec.org Give to make a special gift or become an Ions
member today so we can continue to offer these free programs. So great. Thanks again, everyone. And
Dean, I know now we just have a couple of minutes left, but I’d love to just hear your closing thoughts.
I know you’re mid research, you’ve given a lot of the implications of this, but what does that mean for
you or for any of us watching this, knowing that this research is happening?
Just some wrap up thoughts from you. Well, this is one of half a dozen projects that I’m working on at
various stages. My first two degrees were in electrical engineering. So the engineering side of me is
always looking for something pragmatic, something to make that will do something in the world be
useful in some way. The science side, my doctor is in experimental psychology.
It’s more of an interest in scientific issues like does this exist, how does that work, those kinds of things
which are not necessarily pragmatic and they take years before they turn into something. So this project
is I’m happy about this because it’s a little bit of both. There certainly are scientific issues involved
here. We need to figure out ways of using these kinds of, let’s say, non local consciousness phenomena
that can be immediately obvious to everyone why we would want to study these things. Because
they’re useful, that’s why.
This project, even though it’s at early stages and normally I don’t talk about something where I’m still
involved in working on it, I would normally wait until I had a paper or several papers published to talk
about something with higher confidence. But I figured that people are always interested in how their
own experiences relate to something that actually could be pragmatically useful. And so I thought I
would talk about it for this talk. Absolutely. Absolutely.
Well, thank you, Dean. I know we see a lot of thank yous coming in from the chat. So thank you for
this fascinating discussion. Thank you everyone for posting your thoughts and experiences and
questions. The chat and in the Q and A, we are so grateful and we’ll see you all soon.
Thanks again. Have a great week.