The preoccupation with the future runs the gamut from the season cycles for planting and harvesting, to guessing what civilization and humanity look like in the future. One could say that all our efforts at raising and educating children center on our desire to shape and influence the future.
Throughout the history of humanity, we have sought the counsel of prophets, sibyls, soothsayers, channelers, and visionaries. We have pursued science, which at its heart involves predicting future events by observing cause and effect, and predicting outcomes based on a given set of circumstances.
We all do trending, which simplistically is looking at the historic fact that the Sun has always risen in the east, therefore we predict it will continue to rise in the east, barring unforeseen events. However, it is those unforeseen events that drive our ongoing quest to peer into the future.
What's more, we want to see the future in a repeatable and reliable way. We seek a scientific means to gauge future events, and perhaps even explore optional choices before we make them with some certainly of the outcome. Others would seek to capitalize on knowing the future and placing bets, like the character in "Back to the Future," who had sports scores decades in advance.
In the past decade, a number of efforts have been initiated using what are commonly called 'web bots' to scour the internet and build future scenarios on our use of language. At their heart, they marry two theories together, one being the ancient art of SKED, which seeks to interpret a speaker's or writer's true meaning by gauging the emotional content of his word choice. The other is the proposition that we are all clairvoyant at some level, and that our insights are buried in our daily discourse.
A number of companies, and probably clandestine political operations, have sprung up in recent years that use some form of this 'language parsing' technique in order to predict market trends, gauge the temperature of large groups of people and even guess at the likelihood of a new product's success, much like a massive focus group or survey.
These techniques would have been unthinkable just 20 years ago. Until the advent of the internet, and things such as the very blog you are reading now, there was no easily available sample of mass communication. With the internet, however, there are now literally billions of people all putting forth their ideas and hopes and fears on a daily basis, and because of the digital nature of the information, programs can be written that can amass a large sample of global text and parse it for hidden meanings.
Argueablly, one of the first and best known developers of this technology is Clif High, whose website HalfPastHuman offers frequent updates for a small fee that ostensibly allow the reader to prepare for future events.
Mr. High is a rather private individual who doesn't seek a lot of publicity, however his associate George Ure of UrbanSurvival, has done a fine job of promoting the web bots. A simple search will bring up many articles, statements, posts, and interviews by and concerning Mr. High's work. We sent a few questions to Mr. High, trying to get a little more understanding of this technology and what it promises and what its dangers are. He graciously responded thusly:
LFS: In a nutshell, what are the web bots? How do they work?
CH: spyders scour the net for usual, casual conversations (no emails, no chat, no private or
protected areas sampled). They look for over 100,000 key words. They bring back 2058
characters on either side of the word (when found). These are then processed as 4 digit hex
numbers in Prolog (AI programming language) that aggregates the words around archetypes (family, food, pies, nature, danger....et al). Then i examine a modelspace filled with tens of
millions of words and interpret the changes in language around the archetypes.
LFS: How did the idea to develop this technology come about?
CH: While working for a software firm (largest on the planet at the time) i came up with a new
way to store info on computers which led me to examine the nature of language and
communications, and i discovered what seemed to be trends developing.
LFS: What was the first indication you had that you were on to something with this technology?
CH: 1997 when looking for SUN computing stock info, i came across huge amounts of info
about the sun (big scary solar ball in sky) that associated it with Disease, just before a huge
CME experience. Then i explored further and saw that in 1997 it was bringing back language
about the upcoming , year 2000, stock market internet bubble crash.
LFS: How accurate would you say the web bots are, in terms of trends analysis and predictive power?
CH: that is a matter of opinion and it varies. My routine saying is that we are probably twice as
good as 'random chance' should allow.
LFS: Are you able to view results in real-time? If no, what is the limitation?
CH: No. Too much processing to be done. And too much human interaction that provides most of
the interpretation. So limit is human brain power and time.
LFS: Have you ever encountered similar bots roaming the internet?
CH: yes. 3/three times i have had such encounters with our software. But not since 2006.
LFS: Have you ever traded on the information gathered by the web bots?
CH: Once. in 2000 i used the warning provided to sell all our 44 shares of microsoft stock a mere
8 working days before the crash and one day prior to the all time high of the stock.
LFS: Are you able to parse video/audio recordings at this time? If no, why?
CH: no. Too much work. And no $$ to put into the massive amount of resources (computing)
required to reduce video or audio to text.
LFS: Do you see any ethical considerations with tying this kind of technology to high-speed trading computers?
CH: no, not really as it is the human interpretation that creates the meaning from the mass of
data. So there will always be that hard stop of running into human time as opposed to
machine time.
CH: spyders scour the net for usual, casual conversations (no emails, no chat, no private or
protected areas sampled). They look for over 100,000 key words. They bring back 2058
characters on either side of the word (when found). These are then processed as 4 digit hex
numbers in Prolog (AI programming language) that aggregates the words around archetypes (family, food, pies, nature, danger....et al). Then i examine a modelspace filled with tens of
millions of words and interpret the changes in language around the archetypes.
LFS: How did the idea to develop this technology come about?
CH: While working for a software firm (largest on the planet at the time) i came up with a new
way to store info on computers which led me to examine the nature of language and
communications, and i discovered what seemed to be trends developing.
LFS: What was the first indication you had that you were on to something with this technology?
CH: 1997 when looking for SUN computing stock info, i came across huge amounts of info
about the sun (big scary solar ball in sky) that associated it with Disease, just before a huge
CME experience. Then i explored further and saw that in 1997 it was bringing back language
about the upcoming , year 2000, stock market internet bubble crash.
LFS: How accurate would you say the web bots are, in terms of trends analysis and predictive power?
CH: that is a matter of opinion and it varies. My routine saying is that we are probably twice as
good as 'random chance' should allow.
LFS: Are you able to view results in real-time? If no, what is the limitation?
CH: No. Too much processing to be done. And too much human interaction that provides most of
the interpretation. So limit is human brain power and time.
LFS: Have you ever encountered similar bots roaming the internet?
CH: yes. 3/three times i have had such encounters with our software. But not since 2006.
LFS: Have you ever traded on the information gathered by the web bots?
CH: Once. in 2000 i used the warning provided to sell all our 44 shares of microsoft stock a mere
8 working days before the crash and one day prior to the all time high of the stock.
LFS: Are you able to parse video/audio recordings at this time? If no, why?
CH: no. Too much work. And no $$ to put into the massive amount of resources (computing)
required to reduce video or audio to text.
LFS: Do you see any ethical considerations with tying this kind of technology to high-speed trading computers?
CH: no, not really as it is the human interpretation that creates the meaning from the mass of
data. So there will always be that hard stop of running into human time as opposed to
machine time.
LFS: Have you ever been approached to design such a system (that you are willing to admit)?
CH: Yes, several times.
LFS: In your opinion, what are the limitations, if any, to such technology?
CH: Humans.
LFS: Finally, to what nefarious purposes could this technology be turned? What's the downside?
CH: Fascistic 'pre crime' kind of forecasting.
If it all sounds a bit mysterious, then you are getting the idea. The data returned to Mr. High is loaded into something he calls 'modelspace.' Simplistically, key ideas form kernels around which accrue descriptive terms related to the central idea. For instance, the word 'earthquake' may have words related to time, location, intensity, and emotional impact surrounding it.
Groups of these concepts link together to form 'meta-data' layers, which form overall 'flavors' of future events. One example would be 'Terra,' so that several concepts related to Earth and geology, weather, and so on, are joined at higher levels and interact with other 'meta-data.'
Modelspace can be advanced through linear time to observe how certain concepts evolve, change and interact with each other. Doing this can also show certain ideas increasing in intensity, such as the case with words like 'revolution.'
One of the common problems is that the output of the web bots tends to be somewhat poetic or even akin to reading the I Ching. It requires some interpretation on the part of the reader when looking into the future. However, one often looks at the fulfilled events and says, "Ah hah!" The language is suddenly clear in retrospect.
For instance, a while back there was a lot of language to the effect of a major situation involving 'ships' and 'navy' and 'vessels.' It turned out to be a prediction of the last shuttle disaster, which is surrounded with naval and seafaring language.
George Ure also postulates that similar events that are temporally close together tend to 'bleed through.' For example, language surrounding Katrina and Banda Aceh were jumbled together, since both involved large flooding events and severe emotional states.
For this reason, it's not a perfect technology. Both High and Ure admit problems with nailing certain specifics about events. Things like precise location tend to be difficult, and timing of events is usually a matter of following temporal markers (i.e.- this event will follow that one).
Still, having what appears to be a fairly accurate map of future events, at least in terms of emotional impact and type of event, gives one just a slight advantage over random chance.
High has established a number of personal rules, such as not trying to predict the future of specific individuals, though tests have proved spookily accurate, according to Ure.
The question is whether others will stop at certain ethical boundaries. Like any powerful technology, it is a double-edged sword. Having even a fuzzy view of the future can give certain elements of society undue influence, as High points out.
The other issue is one of destiny. If future events can be accurately predicted, then are we trapped into certain patterns with no hope of avoiding them? Even more interesting to ponder is, if we can predict the future using our language, do we then 'speak' our future into being? And if so, does it not place a grave imperative on what we say?
The concept of the web bots is quite fascinating. Obviously, if companies are springing up that offer similar technology, could an entire industry spring up around future-casting? And does that bring us back to destiny as unavoidable?
The implications are rather astounding. In fact, thinking about them can lead one into circular arguments, not unlike the temporal loops and paradoxes of science fiction. One hears the line from "Spiderman" repeating over and over, "With great power comes great responsibility."
The question then is, does everyone who possesses or will possess this technology feel the weight of responsibility? Or will it become yet another tool in the hands of those who would repress humans for personal gain?
We can hope that people such as Mr. High and Mr. Ure, whose openness with this tool, will set an example for others. If the future belongs to all of us, then we are due any information that can help us shape it. If, indeed, it can be shaped.
The Shape Of Things To Come reports are available through Clif High's website: http://www.halfpasthuman.com, for $10 per report (roughly quarterly). Mr. Ure's website is http://www.urbansurvival.com/week.htm. Massive amounts of information can be found by searching for favorite engine, as well.
No comments:
Post a Comment
Feel free to leave your own view of The Far Side.