Question
How much of the aggregate technological progress of mankind has been achieved during the last century?
Answer
I think you mean Kenya and Tanzania, not Ethiopia, but never mind that.
Your question has a hidden assumption that technology is an objective kind of "progress" in the sense of evolving in a specific direction (and possibly also towards a specific end state), and that progress being a "good thing" in some sense. This type of assumption is called "historicism" and has its critics, so keep that in mind. It has had its major champions as well, but the heyday of this philosophy of science/technology was probably the modernist movement of the 20s. But whether you believe the "progress" is good or bad, you can separate the moral value of the directional trend from the trend itself. So for the rest of this answer, I'll treat "progress" as just a word implying directional, cumulative change, and leave the moral interpretation of whether that is good/bad to you.
Next, you need to define what metric you are implying when you use a quantitative idea like a "percentage." On the face of it, it is an empirical question with many proxy variables for progress (ranging from number of patents/papers published to financial impact of the average invention to per-capita productivity or a standard-of-living bundle, to the maximum speed humans can travel (from walking to space shuttles...) etc.) The question would be very weak if it turned out that some of those variables trended up ("progress") but others did not.
For the record, John Horgan, a popular philosopher of science, believes we are going backwards on many fronts ranging from disease to space travel.
http://www.scientificamerican.co...
My hypothesis is that quite a few variables will trend up, constituting empirical proof of "progress." Moreover, I believe that they will exhibit the same phenomenology (example, linear, polynomial or exponential growth).
But this is still weak, until you come up with a fundamental analytical theory and model of the progress. In other words, it needs to be more than an empirical question, otherwise you're left with the dumb sort of position that "the 20th century accounts for 50% of all progress because it accounts for 50% of all scientific papers and patents."
I believe something like this can be constructed using information-theoretic tools (specifically, an idea called Kolmogorov-Chaitin entropy, which could be applied to all of technology by viewing everything as part of a single giant computer program that is slowly growing more powerful, and calculating that program's theoretical "efficiency"). To my knowledge nobody has done anything like this, though there are complexity theorists who've tried more modest ideas like this. It would be a horrendously difficult theoretical and empirical task to build and validate such a model. But to vastly oversimplify, it would be basically like tracking the evolving "complexity" of a human being from a single fertilized cell to maturity OR like tracking the speciation complexity of an ecosystem. My conjectured theoretical model is somewhere between an ontogenic one (a single organic entity... a theory called autopiesis: http://en.wikipedia.org/wiki/Aut... ) and an evolutionary one, in other words.
If somebody gave me enough money and research assistants, I'd be happy to spend the rest of my life studying this question.
With all those caveats out of the way, what should we expect to see? The historical evidence (qualitative, narrative) suggests that technological progress is an exponential curve at least in the last few centuries, and with adjustments for noise events. Read Joel Mokyr's "The Lever of Riches" to get a sense of the historical data. Mokyr's basic suggestion for the phenomenology underlying the growth is "path dependence"... once one technology is created, others become possible. Path dependence is the explanation he uses to explain why, for instance, despite Europe and China being roughly at the same level of technological development at the start of the Renaissance, Europe leaped ahead. Others have used cultural essentialism, environmental determinism and even religious values as explanations, but path dependence is certainly a major factor, and is potentially an explanation that can subsume the others.
Or to simplify the whole complicated thing down to the kind of model an overconfident math/engineering undergrad might make up, technological progress is a first-order process: dx/dt = kx. The more you know, the faster you learn. The more you've invented, the faster you invent more stuff.
Solve that and you get an exponential. Between that paper napkin hypothesis and actually showing exponential technological growth, there's a ton of careful work to be done, but a lot of people have done/are doing bits and pieces of it, and I'd say it's roughly true. At least so far, during the growth phase of "world civilization." Whether we can grow for ever in the sense of technological knowledge, is another question. There may be limits. Humans do stop growing at some point. Maybe technological ecosystems do as well.
Now to your question of percentage, it obviously depends on the rate of exponential growth, after adjusting for "noise" variables like wars, depressions, plagues etc. My guess would be "50% or more" for the amount of "progress" that has happened in the last 100 years, over the previous 12,000 (the Neolithic revolution of around 10,000 BC, is a better starting point for cumulative technological evolution than the first hominids for a variety of reasons).
Keep in mind that this is a very wildly speculative answer. It's easy enough to shred your question, and my answer, to pieces. To give you an example of the sort of subtlety you have to resolve, could you claim that Newton's laws were the basic technological advance that accounts for all modern mechanical engineering? If not, what percentage of the a new machine design should be thought of as "new" (and therefore "progress") and what percentage a trivial working out of Newton's laws? If you make hammers and start selling a new hammer with a blue handle this year, does that count as "progress"?
There are some potential answers to such difficulties, but let's not go there right now...
Your question has a hidden assumption that technology is an objective kind of "progress" in the sense of evolving in a specific direction (and possibly also towards a specific end state), and that progress being a "good thing" in some sense. This type of assumption is called "historicism" and has its critics, so keep that in mind. It has had its major champions as well, but the heyday of this philosophy of science/technology was probably the modernist movement of the 20s. But whether you believe the "progress" is good or bad, you can separate the moral value of the directional trend from the trend itself. So for the rest of this answer, I'll treat "progress" as just a word implying directional, cumulative change, and leave the moral interpretation of whether that is good/bad to you.
Next, you need to define what metric you are implying when you use a quantitative idea like a "percentage." On the face of it, it is an empirical question with many proxy variables for progress (ranging from number of patents/papers published to financial impact of the average invention to per-capita productivity or a standard-of-living bundle, to the maximum speed humans can travel (from walking to space shuttles...) etc.) The question would be very weak if it turned out that some of those variables trended up ("progress") but others did not.
For the record, John Horgan, a popular philosopher of science, believes we are going backwards on many fronts ranging from disease to space travel.
http://www.scientificamerican.co...
My hypothesis is that quite a few variables will trend up, constituting empirical proof of "progress." Moreover, I believe that they will exhibit the same phenomenology (example, linear, polynomial or exponential growth).
But this is still weak, until you come up with a fundamental analytical theory and model of the progress. In other words, it needs to be more than an empirical question, otherwise you're left with the dumb sort of position that "the 20th century accounts for 50% of all progress because it accounts for 50% of all scientific papers and patents."
I believe something like this can be constructed using information-theoretic tools (specifically, an idea called Kolmogorov-Chaitin entropy, which could be applied to all of technology by viewing everything as part of a single giant computer program that is slowly growing more powerful, and calculating that program's theoretical "efficiency"). To my knowledge nobody has done anything like this, though there are complexity theorists who've tried more modest ideas like this. It would be a horrendously difficult theoretical and empirical task to build and validate such a model. But to vastly oversimplify, it would be basically like tracking the evolving "complexity" of a human being from a single fertilized cell to maturity OR like tracking the speciation complexity of an ecosystem. My conjectured theoretical model is somewhere between an ontogenic one (a single organic entity... a theory called autopiesis: http://en.wikipedia.org/wiki/Aut... ) and an evolutionary one, in other words.
If somebody gave me enough money and research assistants, I'd be happy to spend the rest of my life studying this question.
With all those caveats out of the way, what should we expect to see? The historical evidence (qualitative, narrative) suggests that technological progress is an exponential curve at least in the last few centuries, and with adjustments for noise events. Read Joel Mokyr's "The Lever of Riches" to get a sense of the historical data. Mokyr's basic suggestion for the phenomenology underlying the growth is "path dependence"... once one technology is created, others become possible. Path dependence is the explanation he uses to explain why, for instance, despite Europe and China being roughly at the same level of technological development at the start of the Renaissance, Europe leaped ahead. Others have used cultural essentialism, environmental determinism and even religious values as explanations, but path dependence is certainly a major factor, and is potentially an explanation that can subsume the others.
Or to simplify the whole complicated thing down to the kind of model an overconfident math/engineering undergrad might make up, technological progress is a first-order process: dx/dt = kx. The more you know, the faster you learn. The more you've invented, the faster you invent more stuff.
Solve that and you get an exponential. Between that paper napkin hypothesis and actually showing exponential technological growth, there's a ton of careful work to be done, but a lot of people have done/are doing bits and pieces of it, and I'd say it's roughly true. At least so far, during the growth phase of "world civilization." Whether we can grow for ever in the sense of technological knowledge, is another question. There may be limits. Humans do stop growing at some point. Maybe technological ecosystems do as well.
Now to your question of percentage, it obviously depends on the rate of exponential growth, after adjusting for "noise" variables like wars, depressions, plagues etc. My guess would be "50% or more" for the amount of "progress" that has happened in the last 100 years, over the previous 12,000 (the Neolithic revolution of around 10,000 BC, is a better starting point for cumulative technological evolution than the first hominids for a variety of reasons).
Keep in mind that this is a very wildly speculative answer. It's easy enough to shred your question, and my answer, to pieces. To give you an example of the sort of subtlety you have to resolve, could you claim that Newton's laws were the basic technological advance that accounts for all modern mechanical engineering? If not, what percentage of the a new machine design should be thought of as "new" (and therefore "progress") and what percentage a trivial working out of Newton's laws? If you make hammers and start selling a new hammer with a blue handle this year, does that count as "progress"?
There are some potential answers to such difficulties, but let's not go there right now...