Below is our first guest article and guest speaker in the future conference. As we get more this page will be reorganized.
You are sitting at your homework study center with The Futurist reading a science and technology (S&T) forecast, and wondering what to make of it. Is it credible? Should you pay attention to it? The purpose of this article is to help you become a better consumer of S&T forecasts. It suggests a simple, handy checklist of criteria (shown on page 3) that you can refer to in order to know whether, or how seriously, to take a forecast.
Your use of forecasts may range from becoming more informed about the future, to a hobby or entertainment pursuit, or as part of your professional life. However used, a good forecast is one that stimulates you to think further. It may even challenge your mental model of how the world works. Science and technology forecasts can provide a road map to, and vision of, the future. They will be increasingly useful -- or harmful -- as the importance of science and technology to society continues to grow. The approaching millennium is likely to spur a great proliferation of forecasts as well. The checklist will help arm you for this onslaught.
You should do one of two things with this checklist. One is to take it as is and use it as a handy guide to refer to when reading a forecast. The other is to modify it to suit your specific needs. If you are using it as an input to business planning, you need a certain kind of output. You may be looking for new product ideas and emphasize the capabilities component of the checklist. A forecast of fuel cells as the next major transportation power source may suggest new business opportunities to your automotive or plastics company. If you are in government, you may emphasize implications' component. You might rethink whether to continue spending millions of dollars on new battery technology. If you are reading forecasts for your own intellectual purposes, it is yet another evaluation you make. You may pay close attention to trigger events in order to know what to look for as the technology develops.
The checklist can also help you engage a radical, fringe, or outlier forecast -- one that seems outrageous at first glance. Rather than dismissing it out of hand, subject it to the checklist criteria. If it does not meet them, you might file the forecast away as amusing but reckless. On the other hand, you might consider scenarios under which it could happen, as a way to test your thinking. Could you, for instance, in response to a forecast that Africa will have the highest GDP per capita next century based on a breakthrough in genetics, devise a plausible scenario where the author comes up short?
A strong indicator of a forecaster's reliability or credibility comes from who gives it. Some forecasters merit immediate attention. In telecommunications, anything written by James Martin at IBM or Robert Lucky at AT&T has instant credibility. Some journals, such as the futures publication Technological Forecasting & Social Change or the mainstream Scientific American exercise strong editorial review. Forecasts appearing in journals like these are generally sound.
Japanese science and technology forecasting is solid. Their S&T forecasts are similar in that their typical method or strategy is normative. The forecasts are targets or goals toward which government and industry are going to work. Many of their forecasts read like "we will have a fiber optic network by 2015." You can assume that forecast will be used by government and industry as a goal to which they are going to devote resources.
On the other hand, U.S. government S&T forecasting is uneven and short term. The Congressional Office of Technology Assessment, nominally charged with looking at the future of technology, in practice does a wonderful job of reviewing the state of the art. Similarly, agencies such as the National Research Council and the National Science Foundation do a nice job in setting the table for the future, but do not explore it in much depth.
You should also consider that forecasters are sometimes ill served by their editors. They either misunderstand the topic or seek to boost readability by snazzing up a forecast with a catchy title and extracting out some provocative bits and leaving some crucial points out. For example, a colleague wrote a piece on the future of health and medicine that explored alternative health futures. The editor, without asking him, entitled it "How You Can Live to Be 100." He did not write about that, nor does he believe that he possesses the secret to living 100 years. You may come across titles of forecasts that promise much more than they deliver -- and more than the author intended to deliver.
Another thing to consider is whether the author or institution involved has an agenda. Are they enthusiasts or advocates? Nanotechnology, which has a legitimate long-term potential, does not exist except in the minds of a few pioneers or visionaries. When looking at forecasts of nanotechnology, keep in mind that it has not produced anything yet. But there have been books, volumes, conferences, and articles on the topic. So when you read a forecast of nanotechnology, be wary of the enthusiast or proponent, who may not present a balanced picture of its prospects. The interest may be financial, but more often it is intellectual -- authors may hope to see their prediction proven correct or simply believe the technology is a good thing.
On the other hand, the fact that forecasters have a particular bias or are advocates does not necessarily disqualify them from doing a good forecast. What you have to look at is whether they are being upfront about it. If theforecaster comes right out and says something like "we have been working onthis technology for a long time and we believe the market is now ripe," you can bring a healthy skepticism to your evaluation, i.e., you know it might be overoptimistic. The real difficulty you face is when a bias or advocacy is hidden. The best course in this case, if you have the time and means, is to investigate the author or institution. You can also try to infer from the tone of the article whether it sounds like it someone is trying to sell you something.
Most forecasts do not use formal methods. The formal methodology seems to have receded into history. Professional futurists or forecasters generally use formal methods, however, and disappointingly, they make a relatively small share of forecasts. Most of what you will confront is off-the-cuff speculation by experts or people who work in a specific field or on a particular subject. A cautionary note is that an expert on a subject may not be an expert on its future. A musician may have nothing useful to say about the future of music and an architect may have nothing useful to say about the future of architecture. In many cases, the most interesting things forecast about a field are by people outside it. In an analysis of years of mechanical engineering forecasts by Coates & Jarratt, Inc., for example, the best one was one by a firm outside the industry. It was the one forecast that stepped outside the bounds of a field weighted down by conventional wisdom. It suggested links to other engineering disciplines that were novel insights.
The Delphi is the most commonly used -- and often misused -- technique. A Delphi can sometimes confer a sense of methodological rigor that really is not there. A poorly designed Delphi survey is often dressed up and made to look credible. Its strength in reaching a consensus is tempered by its weakness of neglecting or downplaying the outlier or fringe ideas, which often are the most interesting. It is best used as an input to further thinking and analysis, rather than an output or final product.
The use of formal methods such as Delphi, cross-impact matrices, or scenarios, is more often than not a positive indicator that a forecast is sound. But you can easily be misled by assuming that any use of formalmethodology confers legitimacy.
Assumptions are the basic framework within which a forecast is projected to operate. Coates & Jarratt, Inc. devoted a whole chapter to assumptions in building its forecasts for Project 2025. They can be conveniently broken into technological and social contextual categories. A technological assumption for a genetics forecast might be that the tools will be available for decoding the human genome. A social contextual assumption might be that ethical objections to tampering with human genetics would not block applications.
Assumptions are rarely made explicit by forecasters. The fact that a forecaster states his or her assumptions up front is an indicator of a good forecast. An example could be, "I am assuming that the U.S. economy is going to grow at 2% per year," or "I am assuming that we are going to have an information infrastructure in place." Making assumptions explicit makes it easier to follow the forecast's logic. A forecast that smart cars on smart highways will chauffeur people around, for example, would not jibe with an assumption that the Federal budget deficit is a serious problem and will seriously deplete or bankrupt the treasury. Being explicit also forces the forecaster to confront his or her biases. An author of a forecast of environmental technology that adheres to a deep ecology school of thought should let you know that. That belief may be a key factor in shaping what, if any, environmental technologies the author projects to be adopted.
You will most often have to impute the forecaster's assumptions. That is, in cases where the assumptions are not stated, you deduce what the forecaster must be assuming in order to get from the present to the described future. If a forecast projects the colonization of Mars, for example, some of the things it is assuming is that we will have the launch technology, the ability to terraform the Martian ecosphere or build life-supporting biospheres, the money to pay for it, and people willing to do it. As you go through this process repeatedly, you are likely to become a solid believer in the conclusion that more forecasts go wrong with hidden assumptions than with anything else. What may be a perfectly reasonable forecast often falls apart because it is based on unrealistic or faulty assumptions. Many forecasts of bright prospects for renewable energy technologies, for example, went astray because they assumed that oil prices would rise. They did not explore the possibility that oil prices would remain low, and what that would mean for renewable technologies. Often what is missing is a sense of constraints or a failure to explore potentially negative downstream effects. You may find that if you are confused by a forecast, it can be traced to hidden assumptions.
The lack of a clearly stated time horizon is a common shortcoming of S&T forecasts. You will come across far too many vague and imprecise projections -- "in the next few decades," "over the next generation," or "in the future." The temptation to use these phrases in doing a forecast is sometimes great. The process of putting a specific date on when a forecast is going to occur, however, forces forecasters to clarify their thinking. It helps make the future, an inherently fuzzy and abstract concept, more concrete. A forecast with a specific time horizon helps you to create a mental image or model of the described future and is an indicator that the forecast is well thought out. On the other hand, do not take the dates too literally. They are meant as approximations rather than guarantees.
Here you engage the cognitive content or the core of the forecast. You are looking for the author's specific projections. As you read them, think about the system in which they are embedded. What could enable it? What could block it? Do a little forecast in your own mind. Consider a forecast that intelligent machines or robots will be in widespread use as household servants in 30 years. What else would have to happen for this to occur? Surely, advances in computer power, fuzzy logic, machine vision and mobility would have to continue. Society would have to be affluent enough to afford them. What obstacles could arise? Perhaps an early generation household robot malfunctions and harms or kills a child and they are subsequently banned. Going through this process will help clarify your own thinking. You can also see whether the author addresses the questions that you came up with and weigh his or her response, if any. The other elements of the checklist are essentially an aid to determining whether or not you should pay attention to this core of the forecast. If the other checklist items have been adequately addressed, then the forecast merits serious attention. If they have been underplayed or ignored, you should be skeptical.
Trigger events are developments expected to occur over the course of a forecast intended to let you measure how it is tracking. It is not a typical component of forecasts. You will rarely find a forecast that makes them explicit, but they are occasionally embedded. Some forecasts use a time line approach that lays out developments every five or 10 years that build upon one another.
Consider a hypothetical forecast of the state of the environment in 2025 in which major issues today such as global warming and nuclear waste disposal are being successfully addressed through international cooperation. You would want to know how we got there from here. It would help you decide whether to take the forecast seriously. You might be skeptical of a forecast that suggested extraterrestrials came to Earth and taught everyone how to get along. On the other hand, if a trigger event is the formation an organization called the International Global Warming Federation in the year 2000, you might decide that is plausible.
The trigger event also gives you the means to track the forecast against what actually happens. You can look to see if organization like the International Global Warming Federation -- not necessarily with that title -- is formed. If so, that would be an indication that the forecast is on track. If not, then you may have to pay more attention to alternative futures.
A cautionary note with trigger events is to beware of the mechanical trend extrapolation. That is, if we have some today, we are going to have a little bit more in the future -- in effect a stepladder from today to the future. As Herman Kahn is said to have once observed, this surprise-free scenario is the most unlikely one.
New capabilities are the heart and soul of technology forecasts. One of the things you can do to get the most out of forecasts is to spend some time exploring the capabilities before jumping into specific applications. For instance, a good forecast of information technology would tell you that one of the new capabilities would be the transcendence or smashing of time and space barriers. Think about what that might mean for you, your workplace, or for society. Many forecasters will skip over identifying this capability and get right into applications like virtual communities or distance learning classrooms. This can close off creative thinking too soon. Practicing this art of identifying or extracting capabilities as a consumer of forecasts will hone your futures thinking skills and enable you to get more out of forecasts.
After you read the forecast, consider what the forecast may have missed. Is there a social force that is going to have a big impact that was not dealt with? Is there a competing technology? Is there a new stakeholder or a new actor on the horizon that has not been accounted for? Asking and answering these questions often triggers fresh insights. Many times you may analyze a forecast and find that everything seems to be in place. But after deliberately taking some time away from it and revisiting it, you discover that things are missing. A fresh eye is a valuable evaluative aid.
Discontinuities or wildcards are the single most consistent and glaring omissions. Discontinuities are events that could radically change a forecast. They have a low probability of occurring, but will have a high impact if they do. A forecast that identifies discontinuities probably is a good one.
Another element often missing is a proper accounting for the resources required to make new developments happen. This error is typical of the enthusiast or advocate. A new technology is often technically feasible, but it may not have a market. A rough rule of thumb is that it takes on average 15 to 20 years for a technology to move from an idea to commercial status. The question of who is likely to pay for the technology should be explored. Videotext trials in the 1970s demonstrated the rudimentary technological capabilities of today's information superhighway and interactive applications such as home banking. Videotext forecasts, however, never adequately addressed the question of who would be willing to pay for it, and numerous trials failed. It has taken another twenty years for sufficient interest to develop for interactive services. Be wary of forecasts that skip over the "who pays" question.
The key value of a forecast is in what it means for you. Does a forecast say anything useful to you? Perhaps only half of the job of a forecast is done until its implications are explored. What does it mean to you, an automotive engineer, that speech recognition will soon be available? Will it facilitate smart cars and smart highways? Is it a new business opportunityfor your organization or a new career opportunity for you?
Your questions will be relevant to your particular interests. Will a new technology affect your job or home life, or your children's? Identifying implications is one of the primary benefits of any futures study. In the vast preponderance of S&T forecasts, you will have to extract the implications yourself. Few forecasts take this step. Those that do are providing a useful service and most likely indicate a solid forecast.
In evaluating a forecast, keep in mind that the two most common errors in forecasting are overestimating the speed of the technology's deployment and underestimating the eventual magnitude of its effects. As Asimov said, the difficult thing to forecast is not the car, but the parking problem. A third significant failing is a too-short time horizon. Many forecasts posing as long term are really about the next few years. An indicator of this is a vague time horizon, such as "in the future" or "in the next decade." Sometimes forecasters say they are looking at the next decade, but they are really looking at the next two or three years. A fourth common shortcoming is a lack of imagination. If there is one precious commodity of a forecast, it is imagination. A good forecast inspires you to think more deeply about a subject or sparks new insights. In many cases, trend extrapolations can be a poor substitute for a lack of imagination -- just because we have it today, we will have a little more tomorrow. No imagination.
On the other hand, if you are not careful, you can fall into a trap of focusing too much on what is wrong when evaluating a forecast. Even a forecast with shortcomings or holes may contain useful ideas. Do not get so caught up in applying the checklist criteria that you disregard all imperfectly constructed forecasts. Even a poorly structured forecast is not a waste of time if it has told you some interesting things. Coming away with one solid, new idea is time well spent.
A final point to consider in the overall evaluation is that a good forecast is not necessarily a correct one. Rather, a good forecast is one which stimulates your thinking and leads to subsequent action. Most forecasts are going to be wrong, but they are useful if they have shaken your conception of the future and forced you to clarify your thinking and modify your actions. It may even be the case that a correct forecast is not useful -- if it is just filed away and does not spur action. A provocative forecast, especially one that follows these criteria, is a good forecast, even if the details are wrong.
As you analyze more and more forecasts, certain shortcomings are likely to become apparent. The following items are likely candidates for your wish list.
Greater use of the checklist criteria: Attention to the criteria, in whatever way forecasters choose to interpret or use them, would help. Most forecasts have at least some of them. Far fewer have most.
More forecasts from professional futurists: While the lack of S&T forecasts coming from the futurist community is disappointing, the ones that are done are usually sound. The Institute of Alternative Futures in Virginia, the Futures Group in Connecticut, and the Institute for the Future and Global Business Network in California are some of the futurist firms that produce sound forecasts.
More imagery: People who are visually oriented are not being served well. Very few forecasts use images. A multimedia approach would benefit all of futures research, not just forecasting. Graphical representations of the future would make the field much more accessible. The Futurist has been making some advances in this area. There is generally at least one graphic or picture centerfold per issue that shows pictures of what the author describes with text.
Stating the purpose of the forecast: Is it a new product or market forecast? Is it intended as a solution to a social problem? Being clear about this would aid evaluating it.
Greater attention to social and cultural matters: A significant weakness of S&T forecasts almost across the board is the absence or cursory attention to social and cultural questions. This oversight is the undoing of many a forecast.
In closing, keep in mind the observation of Arthur C. Clarke that when a group of prominent scientists says that something can be done, they are almost certainly right. And when they say that something cannot be done, they are almost certainly wrong.