Create a Bluetooth device that is a small keyboard (and pointing device) with an adjustable clamp that can hold a smartphone.
Turn any smartphone into one with a physical keyboard rigidly attached.
Perhaps hack one together from a Rii Keyboard or Lenovo N5902.
One of the potential problems with enthusiastic consent as a means to prevent rape is that it merely shifts the problem. The rapist must merely psychologically manipulate the victim into giving enthusiastic consent, of creating an illusion of free will (up until the point of rape). Magicians, political candidates, and advertisers do this all the time. Psychological manipulation has always been an essential component of rape; this one additional task perhaps might not pose much of an extra barrier.
The other problem, well noticed, is that a great many people have had such powerful psychological conditioning about sex that it may be impossible for them to utter enthusiastic consent even if they want to.
Annoying is the extremely high sodium content of many canned foods. The point of canning is preservation, obviating the need for salt also as a preservative. Salt can be added to taste by the consumer.
I suspect it is to cover up the bad flavor caused by canning.
What is the largest cross section of a cube? Candidates are the 1-by-sqrt(2) rectangle, the regular hexagon, and the equilateral triangle. It might be none if these.
Along with the square, are these all the "interesting" cross sections? Interesting vaguely defined as extremal in area by a plane moving normally. I suspect the hexagon is minimal in area.
Create a protocol for modifying remote files.
To conserve bandwidth, edits are submitted as a tuple of a hash of the original file and a binary diff.
One possible optimization is to compute hashes with a Merkel hash tree instead of a linear hash, allowing the file hash to be quickly recomputed for edits in the middle of a file. Both sides independently maintain the hash tree. Unfortunately, this does not work if an edit changes the length of a file. Are there better ways? Perhaps possible if we forgo cryptographic security.
We need the hash to avoid merge conflicts, i.e., if two separate clients edit the same original file in different ways.
A client sends a diff but does not receive an acknowledgment. Maybe the server received the diff but the ack got dropped on the way back, or maybe the server never received it. Prevent things from going wrong whichever possibility. Avoid the Two Generals' Problem.
Protocols like this almost certainly already exist. What are they? (Sshfs could use this, if not already.)
Inspired by Google Drive, or the way Google Drive should work. Online "cloud" storage should be a commodity product, with anyone able to implement the protocol, including setting up a server yourself. (But Google does not want it to become a commodity product.)
Goal is to mix and match any editor with any backend remote storage of your choice.
Git, or any version control system, does almost what I want. We need a way to push, declaring this version should be become the head regardless of any merge conflicts. Also, prune old versions.
Replace the punctuation commands in vi (vim, etc), like / for search, commands using only letters a-z. The goal is to create an editor that can be used with very small keyboards, for example smartphones including touchscreens. Perhaps English words as commands, too, as autocorrect has become optimized for typing those.
Replace numbers 0..9 with zabcdefghi (for things like repeat argument).
ESC button is on-screen. Or remap (say) qq to escape in Insert mode (maybe qw inserts literal q, and qu inserts literal qu). But we don't want Pasting arbitrary text into Insert mode to escape.
Partition the surface of the earth according to the nearest continent (Voronoi partition except with source regions instead of points). Which continental partition is the largest?
Create musical playlists of songs robust against the file moving, changing names, internal metadata (ID3 tags) changing, the song getting reencoded in a different format or bitrate.
Use audio fingerprinting now famously used for automatically detecting copyright violations.
Copyright in America strictly prohibits derivative works, with a few exceptions. One exception is parody.
Because of this exception, we have an amazingly flourishing culture of parody derivative works, with a mind-boggling amount of creativity and effort expended in creating these new works, both commercial and non-commercial. This state of the world offers a glimpse of what the world could be like with weaker copyright, especially for derivative works.
The parody exception demonstrates how copyright shapes American culture to its core. It shapes our brand of humor, heavily parody-based. Exceptions and non-exceptions to copyright shape how we think, how we communicate ideas to one another.
This aspect, freedom of speech, is the most important issue regarding copyright reform. It should trump all other concerns. Copyright is one of the most important issues facing American democracy, even though it rarely makes political waves. Change it and change the world.
What important ideas are not being effectively communicated because of copyright?
The Aristocrats joke can be made even more offensive by making it personal, targeting each audience member specifically.
Perhaps the joke intended at a certain subcommunity, mocking specifically those things the subcommunity holds precious.
Subverting the standard delivery of trying to offend "everyone".
The weak Goldbach conjecture, that all odd numbers greater than n can be written as a sum of at most m prime numbers, for which the most difficult statement is n=3 and m=3, is an interesting problem because it can be (and is being) gradually chipped away at with published proofs for larger n or m.
Is there anything "meta" interesting about the sequence of proofs toward smaller and smaller n and m? A measure of "difficulty" of the proof? Can a machine learn how to prove simple then difficult things? Can we somehow extrapolate a sequence of valid proofs to create a new valid proof?
The Goldbach conjecture is tantalizing because a typical large odd number can be expressed as a sum of 3 primes in a huge number of ways, but the conjecture merely asks for the existence of one. Any other conjectures have this feeling of overkill?
Modern democracy was born in secret societies, according to Shachtman's Wired article about a deciphered Oculist manuscript . Therefore, to safeguard democracy, we need either of two things:
More secret societies, with social and legal safeguards to allow and promote their existence. (Possibly these conditions already exist.)
Or, sufficient freedom such that the activities formerly performed only in secret societies can be performed openly, for example, freedom of speech without the threat of social retribution against politically incorrect speech.
Cryptography can help somewhat with the former, for example, something like Google Hangout but free from eavesdropping and attack by subpoena.
However, the problem of selecting who gets permitted in a secret society still remains a social problem. Maybe technology can help: use machine-learning to discover who is beneficial or not to a society.
Interestingly, laws forbidding hazing must therefore strike a balance, on one hand attempting to decrease the harm caused by out-of-control hazing, and on the other hand attempting to decrease the harm caused by democratically beneficial societies being less able to properly select trustable members. Hazing was reportedly rampant in those secret societies where democracy was born.
Mechanisms to make the game more competitive, for example, the reverse-order NFL draft, make individual games more likely to be decided by razor-thin margins, for example by the actions of a referee.
What level of competitiveness maximizes entertainment value?
Future post: equitable pro league.
Consider a CAPTCHA asking users to read and transcribe printed mathematical notation into a language such as LaTEX or Mathematica.
Mathematics is a two dimensional language, so perhaps difficult for OCR. Are there other two-dimensional languages many people would likely know, for which there is a standard "typed" form?
Maybe music, but that seems too easy.
Arabic or Chinese calligraphy, Elian script.
A complete list of all the known solutions to the Riesel problem in a format designed to be easily machine-readable. k*2^n - 1 prime, k <= 509203.
New solutions will probably be posted on http://www.rieselprime.de/Related/Riesel.htm and W. Keller's page.
A complete list of solutions to the Dual Sierpinski problem in a format designed to be easily machine-readable. k + 2^n prime, k <= 78557.
A complete list of known solutions to the Extended Sierpinski Problem in a format designed to be easily machine-readable. k*2^n + 1 prime, 78557 <= k <= 271129.
Someday, search a full rectangular region instead of stopping at the first prime for each multiplier. Why (if true) are some multipliers more "fertile" for primes than others? Drag Racing Proth Primes.
Previously, the original Sierpinski problem.
Being offended is learned behavior, so it can be unlearned, and you should unlearn it. Otherwise, being able to be offended provides a lever for someone to manipulate you based on your feelings rather than rational thought.
Americans are masters at being offended, according to foreign anecdotes.
"You should feel offended by XYZ."
Given a long formal proof of a theorem, find a shorter proof in an automated or semi-automated fashion. On one hand, this seems an easier task than the original proof because you can terminate any automated search when the length exceeds the original, and maybe you can use the original proof as a scaffold to search for shortcuts.
On the other hand, a shorter proof could require even more "creativity" and "imagination" than the original proof.
Inspired by Feit-Thompson theorem. Is there a short proof of Fermat's Last Theorem?
476054*2^1000+3 is a safe prime. However, its corresponding Sophie Germain prime, 476054*2^999+1, is a smooth number plus 1. Is the safe prime really safe? Are there attacks similar to Pohlig-Hellman?
862811 1030355 1121891 1211939 are some odd multipliers, the last one factoring very smoothly to 23*23*29*79.
1487395*2^2001+3
Here are the least primitive roots (generators) of some very large safe primes discovered by others.
f(k,n)=sophie=k*2^n-1; addprimes(sophie); component(znprimroot(2*sophie+1),2)
f(18543637900515,666667)
time = 6,472 ms.
11
f(183027,265440)
time = 2,373 ms.
5
f(607095,176311)
time = 1,464 ms.
7
f(14962863771,140001)
time = 1,116 ms.
5
f(2540041185,114729)
time = 877 ms.
13
f(109433307,66452)
time = 452 ms.
5
Pari/GP (as of version 2.5.0) seems to have some magical code able to compute the least primitive root of a safe prime much faster than doing even 1 modular exponentiation, after asserting the primality of corresponding Sophie Germain prime.
t(k,n,g)=sophie=k*2^n-1; 1==Mod(g,2*sophie+1)^sophie
t(109433307,66452,5)
time = 55,515 ms.
t(2540041185,114729,13)
time = 3min, 26,513 ms.
t(14962863771,140001,5)
time = 5min, 32,705 ms.
t(607095,176311,7)
time = 9min, 43,161 ms.
t(183027,265440,5)
time = 24min, 24,071 ms.
t(18543637900515,666667,11)
time = 2h, 50min, 5,654 ms.
Perhaps there is an application for very time-consuming Diffie-Hellman key exchange.
See also 16384-bit safe primes and A181356 and comments on RFC 2409 and 3526.
Omitting the 70 multipliers known to be difficult (as listed on Wilfrid Keller's page), the remaining exponents for the Sierpinksi problem can be fairly quickly recalculated with Pari/GP.
tough=[2897, 3061, 4847, 5297, 5359, 5897, 7013, 7651, 8423, 10223, 13787, 14027, 16817, 18107, 19249, 20851, 21181, 22699, 24737, 25819, 27653, 27923, 28433, 32161, 33661, 34711, 34999, 36983, 37561, 39079, 39781, 40547, 44131, 44903, 46157, 46187, 46471, 47897, 48833, 49219, 50693, 51917, 52771, 53941, 54001, 54739, 54767, 55459, 59569, 60443, 60541, 62093, 62761, 63017, 64007, 65057, 65567, 67193, 67607, 67759, 69107, 69109, 71417, 71671, 74191, 74269, 75841, 76759, 77899, 78181];
fs(k, nmax)=answer=0;for(n=1, nmax, if(ispseudoprime(1+k*2^n), answer=n;break));answer
mt=[];for(j=0, 78556/2, k=2*j+1;if(!setsearch(tough, k, 0), z=fs(k, 8000);if(z, print(k, " ", z), mt=concat(mt, k))))
time = 4min, 51,630 ms.
[solutions]
? mt
%3 = [78557]
The complete list of solutions may also be useful as a collection of primes of many different sizes.
The Sierpinski problem concerns itself only with odd multipliers because usually the powers of two of an even multiplier can be divided out and added to the exponent. However, the multiplier 2^16 = 65536 is special. The problem of finding a prime number 65536*2^n+1 (n>0) is equivalent to a classic unsolved problem in number theory: are there any Fermat primes larger than 65537?
Unfortunately, the Pohlig-Hellman algorithm makes these primes unsuitable for Diffie Hellman.
Two 1x1x2 building blocks, with their surfaces partitioned into 1x1 squares the obvious way, can be joined in many different ways, touching at a positive integer number of squares. I think 5 ignoring chirality.
In contrast, two 2x2x2 cubes can only join in 3 ways.
Consider a fairly large graph where nodes are all words, and an edge connects two words if they differ by one letter, either by adding or subtracting one letter (EAT - EAST) or changing one letter (LUCK - LOCK). What is the diameter of the largest connected component? What is the most needed new word or words to connect the two largest connected components? Any other interesting questions?
The size of the graph should cause many questions to be computationally challenging. For a greater challenge, also add all pairs of words, allowing edges like FASTER - FASTPER (fast per).
Replace or augment your home furnace and hot water heater with a rack of servers. Normally electric heating is more expensive than gas heating; however, we can recoup the extra cost by renting out the servers for grid computing. Do the costs work out? We need computers that run reliably in less-than-optimal environments (basements). These computers will only run when heat is needed.
To do CAPTCHA in the style of reCAPTCHA, we need a large corpus of text difficult for OCR but easy (easier) for humans. OCR can do printed text quite well, so handwritten text would be a good next challenge.
Children's letters to Santa might be a good, very large corpus of difficult-to-OCR, mostly grammatical, semantically meaningful handwritten text. Instead of just a word, provide an entire sentence or more to transcribe so the human can work with the semantics.
Where do letters addressed "North Pole" go, and have they been kept? There are more letters every year, guaranteeing an adversary cannot exhaustively solve the entire collection.
Any collection of handwritten essays, perhaps from a school, might work.
Instead of strictly rejecting at a threshold value, provide a range over which the rejection probability increases from 0 to 1. This decreases the need for extremely precise measurements near the threshold.
Inspired by wrestling. Wrestlers try to get as close as possible to the maximum of their weight class without going over.
What would American football be like if players had a weight limit?
A talking shark swims among the survivors of the Titanic, politely asking around for Jack, who, unlike in the original ending, has discovered the Mythbusters' method of floating both himself and Rose aboard the board safely out of the cold water.
James Cameron: The script says Jack is going to die, so he's going to die.
A Predator kills Jack just as he is climbing into the rescue ship.
"You killed Kenny!"
Double the dimensions of the image, placing the original pixels at the (even, even) coordinates. Interpolate the (odd, odd)-coordinate pixels from the 4 diagonal neighbors. Then, interpolate the (odd, even) and (even, odd) pixels (a checkerboard) from the orthogonal neighbors. Repeat, doubling the dimensions again.
Here are some experiments generating synthetic noise starting from a random black-and-white array and random gray scale array. It is also possible to start from the checkerboard stage.
Eventually gamma matters, but not taken into account for this.
Applying pnmdepth 2 or 3 yields abstract art.
Can ReCAPTCHA be used by a Tor hidden service without violating anonymity of the server or the user? I think the service must communicate with the reCAPTCHA central verifier via Tor.
This is desirable because Tor services are often the kind wanting to be abused.
Market oxide gemstones, especially Al2O3 (sapphire, corundum) as gemstones that won't be destroyed in a fire like diamond will.
One traditional hypothesized narrative of why men find promiscuous women less desirable is that a non-promiscuous mate increases the assurance that the child really is your biological offspring and not of some other father. (In contrast, a woman does not need to worry about this, hence the asymmetry between how promiscuous females are regarded, "sluts", and promiscuous males, "studs")
However, technologies such as birth control, pregnancy tests, prenatal paternity tests, low-cost and safe abortions, and even legal "technology" such as pre-nuptial agreements, divorces, and child support make it much easier to discover and take action against an unfaithful wife.
These technologies are relatively new, some perhaps only a generation old. Human behavior of mate selection may lag technological developments by a while. However, if it does catch up, we may be in for a future where promiscuity in women is no longer considered shameful, a future so vastly different from the present as to be incomprehensible.
Create a vehicle capable of long distance travel over all terrains, including deep mud.
Kind of like a tank, without the armor or weaponry but better gas mileage. Probably includes a very large gas tank.
Achieve a tremendous feeling of freedom traveling in one: nothing can stop you.
There were two developments in politics by 2012 that seem to be unprecendented in my limited knowledge of political history.
A Republican Congress, so dedicated to preventing Obama reelection, blocks all legislation promoting economic recovery. This is a political ploy that will cost the country trillions in delayed recovery and seems unprecedented in scale and brinkmanship. Is this characterization accurate? Is partisan politics so intense as to prefer to significantly hurt the country rather than reach a compromise unprecedented? This seems to be a significant flaw in the republican (representative) form of government.
Politics became personal, with people taking positions so personally as to break up friendships and even families. And the issues which people took positions -- e.g., national health care -- are often so complicated that no normal person could possibly have confidence that their position actually is better. This seems to be bad on many levels: bad for the country, bad for communities, bad for personal happiness. Is this unprecedented? Will it continue? What caused it?
Derivative work is the way people advance, gaining credibility and respect. People familiar with the base work can evaluate derivative work in context ("this is way better"). In contrast, it's much harder to evaluate a wholly original work.
The most famous community that does this is academia, especially science, with one author confirming or disproving another's work. Or extending. All these are technically derivative works.
Excessive copyright protection against derivative work prevents progress. It also has a sinister aspect: a barrier to entry into the field. The copyright holder of the previous work can withhold derivative work licensing to those it does not want to succeed.
Extract useful work from the aerodynamic forces on an object falling through the atmosphere.
Inspired by planes being ripped apart falling in an uncontrolled spin. Harness that force for good.
Two turbines with vanes that spin in opposite directions.
Mathematics notation for the nth root of x, where n is a complicated expression, is kind of awkward looking, trying to get the large expression to fit within the crook of the radical. Also logarithm base n, especially where n contains superscripts. They both rely on being able to typeset with a smaller font.
Both of these happen to be the inverses of exponentiation. Instead, I think we need operators like ^ with precedence and associativity defined.
Enumerate many examples of things that can go wrong when left unattended for a long while.
Things tend to decay. Things tend to grow. Rare probability events. Information gets lost.
Set the input. Pedal for a while on the stationary bike, powering a computation device. Read the output.
Integer factorization (inspired by the Lehmer bicycle chain sieve, even though it was not human powered).
Chess program.
Need sleep mode (or an OS the boots extremely quickly), very scalable CPU speed depending on how hard the user is pedaling.
When two parties enter into legal battle with neither side having unlimited funds, the winner -- realistically or cynically -- is who can bullshit better (not who is legally correct).
This would be a wonderful scenario for entertainment "inspired by real events" -- a courtroom drama (or comedy?) where nobody is the good guy. Everybody lies and often gets away with it because discovering truth is too expensive.
Perhaps told with an undertone of outrage that the legal system actually works this way.
A battle over money: to heirs, or to end-of-life healthcare providers?
Either way it becomes ugly when against the patient's wishes.
People are very uncomfortable in framing the debate this way, but it is often a significant chunk of change, the entirety of life savings, so there is tremendous incentive for both sides to win. Money will be discussed extremely obliquely. Expect many code words, euphemisms, and reading between the lines: "Death With Dignity".
Consider a dating site funded by donations. In particular, it asks for donations if the site caused a good relationship. In this way, the site gets feedback on how well its matching algorithms are working, e.g., how well the date went. The feedback may also discover certain people are terrible people: no one donates after dating them. Perhaps they lied on there profile, or the dating site did a poor job of inducing them to tell the truth.
A marketing tagline could be asking for people to scale their donations on par with "dollars per happiness" from other sources of entertainment they've paid for recently.
Many secondary schools hold mock elections on election day, e.g., "Kids Voting", to prepare students for voting in the real world.
Somewhat more realistic and relevant would be to include mock attempts at disenfranchisement modeled on the many illegal and barely legal practices attempted or done in the real world.
The kid might get fooled, but it was only pretend. Let it be a lesson for the real world.
Form some constants using the following operations:
+ * - / ^ log root sqrt sqr
On the following base constants:
e pi phi 2pi phi^-1
It should be easy to enumerate all small expressions.
Inspired by a nerdy clock which had pi-o'clock marked between 3 and 4. We'd only be interested in constants between 1 and 13.
The interval between e and pi o'clock is 25 minutes 23 and 55/60 seconds.
Update: answers and Haskell source code.
The first 15 odd prime numbers, namely 3 through 53, form a product just less than 2^64. 10 more (starting from 59) form the next product that fits into 64 bits. The sequence of cardinalities goes 15, 10, 9, 8, 8, 7, 7, 7, 7, 7, 7, 7, 6, 6, 6,... This sequence can be run length encoded as (15,1), (10,1), (9,1), (8,2), (7,7), (6,26), (5,131), (4,1408), but the Haskell computer program runs out of memory (space leak) while calculating the next term (3,???).
The practical use for this sequence is initial primality testing or factorization of a very large number by trial division. First, divide the large number by a product of small primes so that remainder fits into a machine register. Then test that remainder for divisibility by the small primes using fast hardware arithmetic (as opposed to software arbitrary-precision arithmetic). Repeat for the next batch of small primes.
Upon further consideration, something involving GCD would probably be better.
Search for large n-bit primes p for which the factorization of p-1 is known. Large n-bit primes are of the form 2^n-k for small k << 2^n. When written in binary, they begin with many consecutive 1s. This is vaguely in contrast to primes of the form k*2^n+1, which have a lot of zero bits and p-1 is also easy to factor.
First idea is to use the algebraic factorization p-1 = 2^n-a^2 = 4*(2^v+b)*(2^v-b) where v = n/2-1, b = a/2. Test for primality of the factors, and we get primes 2^n-a^2+1 for the following selected (n,a): (4, 0); (8, 10); (16, 42); (32, 3090); (64, 7722); (128, 67290); (256, 559398); (512, 12469218); (768, 21765282); (1024, 2652078); (1280, 103509018); (1536, 374122770); (1792, 73035570); (2048, 152517702).
Previous different attempts, including safe primes which are probably the"right" way to go.
Second idea is to take for granted that all 256-bit numbers are "trivial" to factorize (even 2^256-5823^2 factorizes in 9 minutes without taking advantage of its algebraic structure), so search for numbers of the form p = 1+2^(n-256)*(2^256-i). This gives us primes for the following (n,i): (256, 190); (512, 105); (768, 163); (1024, 229); (1280, 108); (1536, 595); (1792, 195); (2048, 18); (2304, 174); (2560, 2728); (2816, 859); (3072, 601); (3328, 201); (3584, 1318); (3840, 174); (4096, 1933); (4352, 1254); (4608, 2700); (4864, 268); (5120, 420); (5376, 1204); (5632, 2548); (5888, 709); (6144, 796); (6400, 2920); (6656, 865); (6912, 6973); (7168, 1476); (7424, 6969); (7680, 1611); (7936, 333); (8192, 9240); (8448, 2044); (8704, 5355); (8960, 514); (9216, 2299); (9472, 2814); (9728, 1509); (9984, 411); (10240, 6330); (10496, 2098); (10752, 1890); (11008, 3394); (11264, 513); (11520, 3358); (11776, 3183); (12032, 2385); (12288, 8395); (12544, 5380); (12800, 1419); (13056, 2701); (13312, 5628); (13568, 18001); (13824, 4125); (14080, 20134); (14336, 7845); (14592, 5098); (14848, 4018); (15104, 634); (15360, 19186); (15616, 3991); (15872, 471); (16128, 454); (16384, 5563); (16640, 448); (16896, 4158); (17152, 8908); (17408, 3406); (17664, 3690); (17920, 2028); (18176, 5076); (18432, 5131); (18688, 9333); (18944, 330); (19200, 3313); (19456, 5559); (19712, 4816); (19968, 3673).
Increase transparency in government by eliminating penalties for publicly divulging government information, including classified information.
Key is publicly divulging, which distinguishes it from espionage. Operationally, it is different from espionage because for the latter, there is uncertainty whether the enemy knows the information: the same principle as tamper-evident (as opposed to unpickable) locks and safes.
Government transparency is not something that the government can be expected to do well on its own because of too many powerful agents within who would like to keep secrets. There must be an adversarial component. (Previously, FOIA.)
In other words, increase whistleblower protections.
Bradley Manning would be a free man.
Because the T charges more for Charlie Ticket rides compared to Charlie Card, it has a perverse incentive to make the cards difficult to obtain.
This is exactly what is happening: cards now available only at certain hours at certain stations on certain days.
They should just sell them at marginal cost.
In contrast, if cards and ticket fares were the same price, the T would have incentive to encourage card use to save ticket printing and card stock.
If a culture is creating good art, then something might be wrong: talented, creative individuals are being prevented from doing something more productive with their talents. Every engineer knows there is a tremendous amount of art involved in engineering -- creating something.
Jazz, invented by African-Americans in a segregated society.
Also inspired by, why is there so much bad art?
Chemistry has such tremendous combinatorial possibilities and complex interactions that we may never discover some of them even though they may be wonderful. Conditions too narrow to be discovered by chance; too unexpectedly complicated to be discovered on purpose.
Computers probably fail at trying to do full Schödinger equation simulations of many, many electrons.
What information are people trying to make disappear for the purpose of telling a different story to future historians than what actually happened?
Look backward at similar attempts in the past.
A keystroke triggers a Listener to begin a user-interface computation. We want a framework where a subsequent keystroke can abort the computation in progress and restart a new computation. Internal state must remain consistent.
Automatic spell checker, auto complete suggestions, Google style auto search.
Far more frequent than keystrokes is "mouse cursor moved".
Most relevant when a CPU intensive UI gets run on a slow (possibly heavily loaded) computer.
A player turned sports broadcaster or journalist must not talk about, or must explicitly lie about, certain things, for example the extremely widespread performance-enhancing drug use among baseball players of which he was one.
Study carefully how the broadcasting or journalism institutions helped cover-up the truth and perpetuate lies, and apply what is learned to other fields. We must prevent such corruption in other, more important, journalism.
Consider a sports league where a team becomes disbanded for the season when it becomes mathematically eliminated from championship. The players become available for the teams remaining in contention. The championship game probably becomes an all-star game.
An auction is probably the best way, perhaps using a virtual currency which every team is granted the same amount at the beginning of the season. The salary still remains the responsibility of the original team.
Some sort of deterministic web of ownership transfers among teams would be interesting to set up, if possible.
Dig up the endorsements of candidates who turned out to be bad politicians. Don't trust future endorsements by that endorser.
On the theme of predictions.
Consider a UI that uses the mouse to move the mouse cursor, but never uses the mouse buttons. To cause an action at a point, press a key on the keyboard with the mouse cursor hovering over it. Essentially a two-handed mouse with 108 mouse buttons. Avoid the slight shift sometimes caused by pressing a mouse button.
Should provide UI feedback you are hovering over the desired point, so can't rely on clicks only.
Won't work so well on laptops.
Create a quiz, or some other method, that accurately predicts how you will act in situations you haven't encountered, yet, perhaps stressful or unusual situations. The results might surprise you. Are you the person you wish you were?
Accuracy can be measured, though it might take decades.
No doubt this has already been done in some other fields: military, employment.
I'm disappointed the Blackberry form factor for smartphones hasn't become more popular. Physical keyboard with a pointing device seems much more precise and productive.
Is it patents? (But Droid Pro, Palm Pre, Palm Pixi all copied the shape.)
Iphone influence so powerful among consumers that they don't know what they want?
People wanting to mainly to passively consume content -- entertainment -- rather than do anything? (Is this iPhone influence?)
Create a watch with a second hand only, no minute or hour hand. Essentially a very simple continuously running stopwatch (that doesn't stop), for people who need just that function. (They get minute-resolution time from a cell phone, for example.) The simplicity allows it to be constructed more cheaply than a traditional watch, no need for 60:1 gears.
Inspired by a doctor needing to measure pulse.
How much more difficult is intercity travel over just getting around your neighborhood?
How much more difficult is interstate travel over intercity?
How much more difficult is international travel over national?
How much more difficult is interplanetary travel over global?
How much more difficult is interstellar travel over interplanetary?
How much more difficult is intergalactic travel over interstellar?
How much more difficult is travel between galaxy clusters over intergalactic? Between superclusters?
I've yet to see science fiction at the last scale: Supercluster Wars, with armies taking millions of generations to even meet.
Do people really understand how big the leap is from "going to the Moon" to "going to Mars"? I don't.
There seem to be two forms of rape, and using the same word for both clouds the issue.
We actually need to distinguish between three issues. The first is rape from the victim's point of view, a term to be used, for example, to discuss care and recovery for the victim. It is, in essence, a medical (including psychiatric) issue, much like any other psychological or physical injury.
The other two are rape from the perpetrator's point of view. As an analogy, is it an assassination, a premeditated murder, or involuntary manslaughter? From the victim's point of view it's all the same: he or she is dead. But from the perpetrator's point of view, they are vastly different crimes, with different motivations, different punishments, and -- this is important -- different measures needed to prevent each crime.
When discussing crime prevention, the discussion must be from the perpetrator's point of view, just like for all other crimes. I feel we are doing this wrong for rape. Neverthless, considering rape from the perpetrator's point of view seems a taboo topic.
Here are two categories:
I have no good names for these two forms of rape, and maybe that's the problem: not having a names for these things is what's preventing progress.
The first form of rape is more famous: control fetish rape. The perpetrator is motivated by, and takes pleasure in, making the victim do something against his or her will. The motivation is the feeling of power and accomplishment of subjugating someone else to your will. The motivation is not for orgasm.
I struggle to come up with a good name for the second form of rape: sex fetish rape. Normally the word "fetish" is used for sexual pleasure in something not related to sex, but in this case, it is tautologically taking sexual pleasure in taking sexual pleasure. The perpetrator is motivated by the biological desire to have sex, an urge so strong that it overpowers respect and property lines.
Are there other forms? Is it a gray scale? Where should crime prevention efforts be concentrated for greatest effect? What social or biological forces produce perpetrators of each category?
Two items from George Washington's Farewell Address, avoid international entanglements and avoid political parties, which we have terribly failed on, are game theoretically equivalent.
If everyone else joins a political party, or forms international alliances, your optimal strategy is to do the same. The world might be a better place if no one did, but you can't control other players' actions: Prisoners' Dilemma.
We are so awesome. You are so awesome. Let's merge.
This never happens, signifying a flaw.
Nationalism. Arranged marriages between royalties in olden days.
