top of page







My name is Greg Kolodziejzyk, and I’d like to welcome you to the Time-Machine web app! Time-Machine combines artificial intelligence with my own time tested method to make intuitive predictions about future events. Read more about my research here.


How does Time-Machine predict the future?


Imagine you're trying to guess the outcome of a coin flip tomorrow: heads or tails. 


Instead of guessing the flip directly, you use pictures to help you make your prediction.


  • If the coin is going to land on heads, there will be a picture of a juicy hamburger shown to you.


  • If it's going to land on tails, there will be a picture of a cold ice cream cone shown to you.


Now, your challenge is to “see” or predict which picture will be shown to you tomorrow. And of course, the picture that is shown to you is based on the coin's outcome. If you believe you “see” the hamburger in your mind, you'll predict the coin will be heads. If you "see" the ice cream, you'll predict tails.


This idea of making intuitive predictions by “viewing” or focusing on associated objects or images, rather than the actual event or target itself is called “Associative Remote Viewing”. Scientific experiments have shown that there is a statistically significant effect, but it is VERY VERY small. The Time-Machine app uses repetition to form a consensus, along with some proprietary artificial intelligence technology that significantly increases confidence in your predictions.


What kinds of predictions does Time-Machine make?


My company is sort of like a hedge fund, and one of many steps in our investment decision making process is the consideration of intuitive predictions. The Time-Machine web app was developed to allow us to crowdsource these intuitive predictions about various investment opportunities.


How do you ‘see’ something you haven’t seen yet?


The process of “remote viewing” is personal, and can vary from person to person. Generally, I find it most useful to close my eyes and relax, reaching a meditative state. Then I imagine that I am looking at an image on my computer screen in the future after the predicted event has transpired. Since it is impossible to know anything about the image you will be shown, as they are chosen from a library of over 20,000 random images, your brain can’t become involved in the process, and you have to listen to your subconscious mind. I try to clear my mind so that I’m not thinking of anything at all, and imaging looking around an image on my screen, waiting for some random thought to enter my mind. 


Most often these random thoughts are generated by your thinking brain, and you must ‘edit’ these out (ignore them), as they are typically thoughts about the last image you saw, your lunch, or anything else that might seem ‘suspicious’ to you. The valuable impressions always come as a big surprise where you wonder how you ever could have generated that thought which seemingly came from nowhere, or ‘out of the blue’. Often, these valid impressions about the photo are very weak thoughts that seem to be occurring in the ‘background’ of your mind, and you might come to the realization that, for example, you had been thinking about a clown's face all the while, while trying hard to think about nothing. Pay attention to these weak random background thoughts, as they are often related to your target.


Please keep in mind that the amount of concentrated, focused effort you invest into a good quality remote viewing session is directly correlated to a successful trial. Taking your time to get into a meditative state is highly recommended. Sometimes I listen to Hemi-Sync sound to help me relax and clear my mind. Search for “Hemi-Sync” on Spotify.


Is there scientific evidence that remote viewing is real?


Aside from my own research, YES! According to a META study conducted by Dr. Dean Radin, a summary of all remote viewing experiments conducted by universities and research organizations, showed an astounding overall effect size (objective method of evaluating experiment results as compared to what could be expected by chance) that resulted in odds against chance of ten million to one. Some individual studies actually resulted in odds against chance of over 100 billion to one. The META study summarized the results of over 5.5 million individual trials that spanned the last century.


According to Professor Jessica Utts, a statistician from the University of California, in a report assessing the statistical evidence for psychic functioning requested by Congress and the CIA, “It is clear to this author that anomalous cognition is possible and has been demonstrated. This conclusion is not based on belief, but rather on commonly accepted scientific criteria. The phenomenon has been replicated in a number of forms across laboratories and cultures.” And regarding continuing research, she adds “I believe that it would be wasteful of valuable resources to continue to look for proof. No one who has examined all of the data across laboratories, taken as a collective whole, has been able to suggest methodological or statistical problems to explain the ever-increasing and consistent results to date.”


I’ve been involved as a researcher in this field for over 30 years, and I know many of the top researchers personally as well as many of the U.S. military remote viewers (various U.S. intelligence agencies explored using remote viewing for intelligence gathering in a 20-year, $20 million basic research program). These are upstanding, highly educated individuals who are motivated to further our collective understanding of this mysterious, yet fascinating field. Read this link for a review of all remote viewing research including the US government programs.


Unfortunately, despite all of the scientific evidence, skeptics still refer to remote viewing as a pseudoscience. Throughout history, there have been numerous instances where widely held beliefs were later overturned by more evidence or new discoveries. Instances such as widespread belief that the world was flat remind us that scientific understanding is ever-evolving. What we accept as truth today might be refined or even overturned as new evidence comes to light.

Is there scientific evidence that time-machine actually works?

YES! Although the percentage win rate of trials so far is small at 51% (51% of trials were correct predictions), this results in a statistically significant z score of 2.8 over 15,527 trials as of Feb 25th, 2024 (z = 1.8 is considered chance expectation). This converts to odds against chance of 390 to 1 meaning that in order for chance to be considered an explanation for our success, we would have to repeat this entire 15,527 trial experiment over 390 times which would take about 130 years! When we combine trials to take advantage of consensus, our prediction win rate jumps to 54%, and even higher when we add confidence filtering.

Screen Shot 2024-02-25 at 1.34.13 PM.png


Managing your time with predictions



Your first prediction

Using the terminology of the Time-Machine app, a prediction is a single “yes” or “no” answerable question about a market event. An example of a prediction is “Will Crude Oil close higher from 10/9/23 02:00 to 10/10/23 01:00 ?” The prediction question is asked 10 times in the form of “10 trials”. Each trial contains two random photos and each of these 2 photos is randomly associated with either “yes” or “no” answers. Your job is to perceive aspects of the single photo that you will be shown after the market event has completed. You repeat this 10 times, once for each trial and Time-Machine will calculate a consensus prediction which is either “yes” or “no”.


When the future event end time has been reached, you must view each of the 10 photos in each of the 10 trials to complete the loop. This is called “feedback”.


When you select “+ New Prediction” in the app, the prediction start time will usually be within the next hour as long as you still have at least 30 minutes to complete all of the trials. The prediction ending time will be from 2, to 12 hours later. You cannot start a new prediction until you have viewed the feedback from any previous prediction.


For example if you start a new prediction at 10:45, the prediction start date will probably be 12:00 because you require 30 minutes to remote view all 10 trials. If the prediction start time is 12:00, then the prediction end time could be as soon as 13:00 if the event is a 1-hour event. If the event is a 6 hour event, then the prediction end time will be 18:00. You can view your feedback any time AFTER the event end time. Keep this in mind if you intend to complete one prediction per day. Please do not make more than one prediction per day.


Since financial markets are closed from Friday afternoon until Sunday afternoon, any new predictions started on Friday will have Sunday at 4:00 pm as the minimum start date.


Recording your remote viewing transcript


Typically, I find that entering a few different thoughts/perceptions formed as short sentences or single words works best. Please don’t be too vague, but also try not to be too specific. The A.I. that is comparing your transcript to the photos might return a low score if you submitted “Checkers. Checkered pattern” and the photo was a black and white photo. In that case, it would have been better to enter “Checkers. Checkered pattern. Black and White”. Don’t be too certain that you perceived EXACTLY what you perceived. Sometimes what you perceive is RELATED to the photo, but not EXACTLY the photo. The question you need to consider is “what was the part of your perception that was actually connected to the target?”. In the black and white example, it was simply the black and white concept, NOT the checkers board.


Clicking the link “Showcase” at the top of the app provides some good examples of high scoring “hits” along with the RV transcript submitted.


 If your primary language is not english, then please refer to a translation app before saving your trial in english.


How the confidence scores are calculated


There are 2 types of scores:


  1. Intended: This is a z-score which reflects how close your RV transcript came to describing a photo in a trial.

  2. Displacement: This is a z-score which reflects how close your RV transcript came to describing, on average, ALL OF THE OTHER photos aside from the intended photo in ALL of the trials for one prediction.


Performance graphs:


Performance graphs show EITHER intended scores OR displacement scores, but NOT any combination of intended and displaced. You can view performance for trials, or predictions. Example: You completed 5 predictions consisting of 10 trials each. If you selected “trials” then you would see results based on 50 trials. If you selected “Predictions”, then you would see results based on 5 predictions. You can also choose to view performance results as % correct values rather than statistical z-scores. You can also isolate specific predictions by entering the prediction numbers into the field “Include only these prediction numbers:”. Example: “1,3,4” will show results for predictions 1, 2, and 4 only. Select “Advanced Settings” to access these options.


My Predictions table:


The table shown on the main page of the Time-machine app is a list of all of your predictions both completed and in-progress. Each line of this table shows the number of trials completed for the prediction, the confidence z-score for a “YES” outcome and a confidence z-score for a “NO” outcome, as well as the actual outcome of the future event (assuming that the prediction is finished).


The “YES” and “NO” z-scores are combined intended z scores + displacement z-score for all trials. For this combined z-score, we compute a Stouffer Z by adding your intended z-score + your displacement z-score and then dividing by sqrt(2).


The following example shows scores for a single prediction:


Trial#   YES z-score   NO z-score

1        1.2           -.5

2        .35           .65

3        -.87          .1

4        .05           -1.4


Total YES Stouffer z = sum(1.2,.35,-.87,.05)/sqrt(count) = .365

Total NO Stouffer z = sum(-.5,.65,.1,-1.4)/sqrt(count) = -.575


The final prediction is the greater of the two Stouffer z-scores, so in this case the prediction is YES. If the actual answer was YES, then this prediction would be marked with a green check “correct”.

bottom of page