That is *great* phrasing - thank you. It's going into my notes and will probably make it into my writings (with attribution). Probably as a chapter heading.

The situation is not completely hopeless: there is a small number of people, myself included, who are working on actual AI. Most of the research is using programming to solve a (particular) problem.

The Turing Test has flaws.

Firstly, it requires a human-level of communication. One cannot use the it to determine whether a crow (for example, or cat or octopus) is intelligent since they cannot communicate at our level. Even though these creatures demonstrate a surprising level of intelligence. Watch this video and be astonished.

The extended video shows the crow taking the worm to it's nest, then returning to grab the hooked wire and taking that back to the nest! Can we use the Turing Test to determine whether the crow is intelligent?

Secondly, it conflates intelligence with human intelligence. There's no spectrum of measurement, no "ruler" which can be laid down to measure the level of intelligence in an entity, or to determine whether one entity is more (or less) intelligent than another. Are crows more intelligent than cats? Can the question be resolved using the test? Could the test be used to determine which of two humans is the more intelligent?

But most importantly, the Turing Test has no predictive value: it cannot be used to guide research or development of intelligence.

Consider trying to build a fizzbin, and whether you are successful will be determined by a yes/no decision from a jury of professionals. With no description of what a fizzbin actually is, how hard would it be?

Consider trying to deliver a package, given that you have a GPS system with a broken display. The GPS still works, and the LED will light when you are at the delivery address, but otherwise you have no idea where to go. The address could be in NYC or Tokyo, or anywhere else.

The fundamental problem with the Turing Test is that it doesn't define intelligence(**). Defining something as a test works in mathematics where there is no time or effort to make the axiom of choice on the set of all objects (ie - the universe), but intelligence isn't a purely mathematical concept. It's partly based on a real-world measurement (being: information), and as such is more closely akin to physics.

Instead of a fizzbin, consider trying to build a car. A car can be defined as a body, frame, 4 wheels, engine, and seats, and the purpose is to transport people from place to place (*). A wheel can be further described as a tire on a rim with brakes, a tire can be described as a loop of rubber with steel wires and a valve-stem, a valve-stem as a tube with a schrader valve, a schrader valve is... and so on.

This is a constructive definition: an object is made of simpler objects, each of which is composed of even simpler objects. Math is full of these (a field is a ring plus some stuff, a ring is a group plus some stuff, a group is a set plus some stuff... and so on.)

With the constructive definition, one could build a car directly; or at least, know how to make the attempt. You can determine whether something is a car; and if not, know what needs to be changed.

In my opinion (I'm an AI researcher) the Turing test and the Lovelace test have little value. The tests don't show where to look or how to proceed.

(*) A simplified definition to not lose sight of the position.

(**) This is an academic position. I am a great admirer of Alan Turing and his many brilliant results, including the Turing Test.

That's an interesting response. You are supporting your position by emotional strength - essentially saying that the poster has to back down or you'll respond into a full-blown emotional outburst (see bully).

When I first heard about Uber's plans the first thing that came to mind is "there's no law against publishing public information".

We have fairly clear rules about what's illegal in terms of gathering and publishing data. The police have no qualms about publishing names and addresses, and sometimes courteously withhold that information for the rich and powerful while using it against low-income people.

The press has no qualms about publishing data that people want to keep private, so long as publishing it would sell papers. If someone simply wishes to live out of the public eye, it's a challenge and "Look! We've got the scoop on Satoshi Nakamoto! Find out who he *really* is and why he needs to hide! (Are your children safe?)

If no one takes action to expose the journalists, if there's no consequences for their actions, what keeps the journalists honest? What incentive does any journalist have for journalistic integrity?

This seems like a cromulent quid-pro-quot. So long as no laws are broken, I'm fine with it.

Your book "Talent is Overrated" is misquoted and misinterpreted in many places, but seems to say that anyone can become a world-class expert with enough effort and time.

What should someone do to become a world-renowned expert?

Can you give us a plan or list of steps to take - something that's not garbled by news media reporting?

Can you clarify a summary of the books conclusions, so that others can embark on that journey?

Part of explaining something is knowing your audience.

Telling someone to type a command in R doesn't explain *why* typing that command works, or what's going on in the background.

And yes, there's things wrong with the post, but Bennett is most definitely NOT A STATISTICIAN. You don't saturate a beginner with all the gory details - you start from the basics and work up.

Part of explaining something is knowing your audience. Practice explaining things to people and you, too, will figure that out.

38 out of 54 survey-takers, or 70%

Bennett, try this experiment.

Make a program that flips 54 coins and notes the number of heads and the number of tails at each round. Then run this program for one million rounds.

When you're done, note the number of rounds the random generator saw 38 or more heads and frame this as a proportion; ie - "the random generator reached this level X% of the time".

Then compare your results with the random generator. If your results are unlikely to come from the random generator, then perhaps you have something.

Now, " unlikely" is an arbitrary measure with no compelling foundation (it's the wrong measure to determine the significance of a result(*)), but in scientific circles we use a "rule of thumb": results are considered significant when they are less likely than 95% of the random results.

Even at this level, we expect 1-in-20 studies to be due to random chance, but then follow-on studies should confirm or deny the findings (and 1-in-20x20 of *those* will be due to random chance as well).

If the results might lead to potentially catastrophic decisions we might use a higher level of significance; for example, 99% confidence when deciding whether a drug is safe. Physics uses an insanely high level of confidence.

Try that and get back to us - we await your next post with baited breath.

(*) The correct measure is the number of bits saved by compressing the original data by factoring out the result (glossing over some details).

I dunno... long-term reading of this blog might result in the impression that life is a disheartening, unjust affair. It's full of rights violations by police and government agencies, feckless and obstructive politicians, corrupt and predatory corporations, and so on.

To read online news results, everything is lurid and emotional. For example, the nurse in Main [who was in contact with ebola] who didn't agree to a quarrantine was in a "standoff" with authorities, the Philae lander is "racing against time" (whatever *that* means), there's a tiger loose in Disneyland, and we need to be afraid of everything so that the government can justify their purchases and policies.

Is it that much of a stretch to believe that people will view the world through this skewed perspective?

Given what we know about human psychology - for example, that people will believe what they're told by default (viz. religion) - it makes perfectly rational sense that a small cadre would lose all hope in humanity and seek to avoid it.

I don't think these people can be legitimately called insane. They're not hurting anyone, they're not hurting themselves, and they're living their own lives.

What criteria would you apply to these people to designate them as "insane", and what behaviour would you change about them to fix it? (And how do measure such a change so that you can tell when they're no longer insane?)

Crap! He promised he wouldn't tell anyone.

Oh well, I guess the cat is out of the bag.

How are people liking the New Age? Any suggestions for improvement?

There was a recent post asking how authorities might have breached the Tor network.

A related question to ask might be "what can we do to increase our network privacy/security"?

I've often wondered if a "government authority" blacklist would be worthwhile. For example, the City Police near where you live probably surf from a fixed IP address at that location. We could maintain a list of such addresses and allow websites to subscribe to the list.

If an address geolocates to within 50 miles of Washington DC (or Langley, VA; or Bluffdale, UT) it's probably not someone you want looking at your site.

Anyone with the slightest idea of how the internet works will realize that this scheme will be trivial to get around using any number of techniques, but the purpose isn't to make access *impossible*, it's to make access *harder*. It starts an arms race between government agencies and an army of determined hackers.

Suppose you're a government agent. You can't send a link over E-mail to your boss at the office because when he opens it the site will show different results. You have to do screenshots or make web page copies - it's much more work (and a more complicated evidence chain).

Suppose you're a government IT guy. You have to implement VPN connections to remote computers so that your agents can surf the net properly, and this is a ton more work for you to do, and it's insecure and might open up your internal network to hackers.

It starts a competition for resources. In addition to law enforcement, the government entities also have to spend time, effort, and money to get around the additional hurdle. If it costs us little to implement, and costs them a lot to get around, then it's effort well spent. And there's a multiplication factor: each and every government agency has to implement a solution to our one system.

In the manner of spam blacklists, we could allow people to nominate specific IP addresses as being "city hall in Tallahassee" with some confirmation protocols to ensure accuracy and that the list doesn't get spammed.

You could have your website either block the listed IP addresses, or show different content.

We could make it *much* harder for authorities to gather website evidence.

As I understand the Tor process, every tine I fire up Tor it randomly chooses an exit node(*).

Suppose I am running some exit nodes (as the NSA is suspected of doing). If I want to find the location of a hidden service I just fire up Tor and access an onion website with a specific tempo. If one of my exit nodes shows traffic with that tempo, then I know that's the exit node for this onion connection and I can trace the exit connection(**).

If you access the site many times, eventually the statistical nature of the tempo (in your own exit node) will be apparent among the random noise of other traffic. If you do the process many times, eventually you'll find a strong statistical evidence for the target IP address.

How many Tor exit nodes does the FBI run? How much time can they put into discovering each site? Can tempo-based access be automated?

See here for more info. From a paper published in 2011 comes the quote:

In this thesis we tested three correlation algorithms. [...] We found that while the two previously-existing algorithms we tested both have problems that prevent them being used in certain cases, our algorithm works reliably on all types of data.

This would be my guess.

(*) For the onion protocol it's listed as a rendezvous point and there's some protocol negotiation, but it's essentially an exit node.

(**) Actually it's even simpler. Tor reports the IP address of your exit node - just keep starting Tor until the exit node is a system you control.

I disagree.

Wikileaks was effectively stopped when all credit card companies refused service. Defense distributed lost their payment processor ("Stripe").

The TOS for many online resellers restrict what you can and cannot sell - eBay won't let you sell booze or their empty, collectible containers, animals, or event tickets. (Why can't I resell my event ticket if I decide I'm not going to use it?) Amazon, even Craigslist have similar restrictions. You can't sell fart apps on the apple store.

This will also put a crimp in the way Corporate Law Enforcement operates. Instead of spending time tracking down the distributor of pirated works, they'll have to fall back to investigating murders, thefts, and assaults.

And then there's the economic upheaval which will happen when previously banned markets become easily accessible. Drugs come to mind, but this will also have an effect on easily-copied data streams such as games, movies, and books. Knowing that your movie will be immediately copied and that you will get no revenue *after* it's made, entertainment might have to switch to a kickstarter-style model. Stephen king proposes a new book, gets $100,000 in seed money, writes it and sets it free on the internet. That sort of thing.

These are just the first few things that come to mind. Some are speculative, but others are happening right now.

I'm pretty sure you're under-estimating the effect that secure untraceable commerce would have on the world.

I had a related discussion with some friends recently about what they would/wouldn't work on in their job.

Einstein and others famously regretted developing the atomic bomb.

At the time, it was thought that nuclear chain reactions were impossible because the neutrons emitted by a fissile nucleus were too fast to interact with neighboring atoms. Leó Szilárd discovered that graphite would act as a neutron moderator, slowing them down so that they could interact. Each decaying nucleus releases two(*) neutrons, each neutron causes two other nuclei to decay, and so on. Two becomes four, becomes eight, in an exponential manner.

Here's the thing. At the time, conventional wisdom felt that chain reactions were impossible; and entrenched ideas in science are hard to pry loose. If Szilárd had chosen not to publish, it would have delayed nuclear fission research for decades - possibly indefinitely.

Consider the ramifications of having a few decades of technological development before attempting to build nuclear reactors, of social development before ICBMs and Mutually Assured Destruction, and so on. We've come a long way since then - we're much closer to planetary cooperation. The conflicts of the early 20th century seem almost tribal in retrospect.

Here's the essential question: Should Szilárd have published? Knowing that his research was the keystone for nuclear weapons, should he have just kept quiet about it?

The tools make no political judgments, but unenlightened bureaucrats do. And right now there's a lot of abuse by the people in power, the people we should be able to trust with our welfare. One only has to look at elections to see how psychological research is being used - en mass - on the population for political ideology.

Would it not be better to put this research off a couple of decades so that other, more directly beneficial technologies can come first? An environment of secure communications, anonymous surfing, safe and untraceable whistle-blowing seems to be on the horizon.

We have the hindsight to see the results of Szilárd's choice. Should we choose differently?

(*) Average 2.5 neutrons per nucleus

To misquote Ash: Republican? Democrat? I'm the one with the vote!

People believe the promises, so the election becomes a competition to see who can promise the best.

Is your life any better for having these party affiliations? Since the last election, has the government made the country better or worse? Will your kids have a harder or easier time when they go out into the world? Will you retire in ease or hardship?

Since the last election, do you have more freedoms or less?

Don't buy into the promises, they mean nothing. Vote against the people in office. That's the way to promote change, that's the way to force people to action.

Vote out the incumbents.

To misquote Ash: Republican? Democrat? I'm the one with the vote!

None of what you said matters - not of the republican words, not democratic promises, not adverts or sound-bites.

What matters is what they've *done* while in office. It's the only metric that matters.

Is your life better since the last election? Will your kids be better off or worse off when they leave the nest to go out on their own? Is the government giving you more freedoms or less?

You shouldn't care which party that is. Vote for change, not for words.

When they realize that they have to actually *do* something during their term in order to get re-elected, then we'll start to see some changes.