Partly-baked ideas

lukeprog

I.J. Good, from the opening of 1962's The Scientist Speculates (a collection of partly-baked ideas):

A partly-baked idea or PBI is either a speculation, a question of some novelty, a suggestion for a novel experiment, a stimulating analogy, or (rarely) a classification. It has a bakedness of p that is less than unity, or even negative. The bakedness of an idea should be judged by its potential value, the chance that it can be completely baked, its originality, interest, stimulation, conciseness, lucidity, and liveliness. It is often better to be stimulating and wrong than boring and right.

A very rough guide to the maximum length that a PBI should have is given by the formula

10^(9px/2) words

where x, the importance of the topic, is between 0 and 1. For example, the maximum length for a negatively-baked idea is less than one word. An idea can compensate in importance what it lacks in bakedness, and conversely. The formula is applicable to each sentence and to each paragraph, as well as to the whole of a contribution. For the non-specialist, the formula makes sense even when px = 1, but in this anthology px rarely exceeds 7/9.

A possible justification for the exponential or antilogarithmic form is that if an idea is developed to a certain length d, then the size of the expository tree increases roughly exponentially with d, if the multifurcation of the tree is the same at every level.

(Note that I changed the formatting a bit for readability.)

I.J. Good, from the opening of 1962's The Scientist Speculates (a collection of partly-baked ideas):

A partly-baked idea or PBI is either a speculation, a question of some novelty, a suggestion for a novel experiment, a stimulating analogy, or (rarely) a classification. It has a bakedness of p that is less than unity, or even negative. The bakedness of an idea should be judged by its potential value, the chance that it can be completely baked, its originality, interest, stimulation, conciseness, lucidity, and liveliness. It is often better to be stimulating and wrong than boring and right.

A very rough guide to the maximum length that a PBI should have is given by the formula

10^(9px/2) words

where x, the importance of the topic, is between 0 and 1. For example, the maximum length for a negatively-baked idea is less than one word. An idea can compensate in importance what it lacks in bakedness, and conversely. The formula is applicable to each sentence and to each paragraph, as well as to the whole of a contribution. For the non-specialist, the formula makes sense even when px = 1, but in this anthology px rarely exceeds 7/9.

A possible justification for the exponential or antilogarithmic form is that if an idea is developed to a certain length d, then the size of the expository tree increases roughly exponentially with d, if the multifurcation of the tree is the same at every level.

(Note that I changed the formatting a bit for readability.)

The book also contains another statement of Good's intelligence explosion idea, later than the original 1959 statement but earlier than the famous 1965 statement (p. 194):

[After bringing a computer to near-human-level intelligence...] We could then educate it and teach it its own construction and ask it to design a far more economical and larger machine. At this stage there would unquestionably be an explosive development in science, and it would be possible to let the machines tackle all the most difficult problems of science.

He then discusses which scientific problems they could solve, what advantages over humans they would have, etc.

He also says "my guess of when all this will come to pass is 1978, and the cost of $10^(8.7 ± 1.0)."