Hopefully Anonymous

How Likely Is It That A Subset Of Humans That Can Pass A Turing Test Aren't Conscious?

How likely is it that a subset of the 6 billion people that can pass a turing test aren't conscious? That they don't have subjective conscious, but are fooling those of us that do? I suspect the odds reasonably would be greater than zero -and perhaps significantly greater. This is the flip side of the argument "if a computer program or an upload can pass a turing test, then it's subjectively conscious". Perhaps some of the 'people' we thnk are subjectively conscious because they pass our current, informal turing tests are not.

Thoughts?

Note: 4/27/08 I just corrected the wording in the title and post to correct a typo that's been in this post for about 6 months. Of course I meant "that can pass a turing test aren't conscious", not "are turing machines". This was caught early in the comments, and I think the rest of the post made it clear what I meant, so I was in no rush to correct it.

The two most important things: (1) Curing Aging, (2) Minimizing Existential Risk

The two most important things: (1) Curing Aging, (2) Minimizing Existential Risk

Links to read de Gray and Bostrom's writings, respectively, on these two topics:

http://www.amazon.com/gp/reader/0312367066/ref=sib_dp_bod_ex/102-0786446-4004163?ie=UTF8&p=S00J#reader-link

http://www.nickbostrom.com/existential/risks.html

A great essay by Arnold Kling on Aubrey de Grey's bestselling book, "Ending Aging".

A great essay by Arnold Kling on Aubrey de Grey's bestselling book, "Ending Aging".

http://www.tcsdaily.com/article.aspx?id=091207G

"Four years ago, I reported that de Grey foresees a not-too-distant future in which humans can reverse the effects of aging, raising the possibility of living healthy lives for hundreds of years. He has not backed away from that position, and this book, written by de Grey and his research associate Michael Rae, represents an update from his perspective. In brief, he says that

1. The latest scientific research indicates no flaws in the theory that aging can be eradicated.

2. However, getting the required techniques developed will require institutional changes relative to our current system for conducting medical research.

As an economist, I am most interested--and most qualified to form an opinion about--the second point."

Arnold Kling's impressive bio:

http://www.tcsdaily.com/Authors.aspx?id=171

"Contributing Editor, TCS

Arnold Kling is a TCS Contributing Editor and an adjunct scholar with the Cato Institute.

He holds a Ph.D. in economics from the Massachusetts Institute of Technology. He was an economist with the Federal Reserve Board and later with Freddie Mac. In 1994, he founded Homefair.com, one of the first commercial sites on the World Wide Web. After Homefair was sold, he wrote "Under the Radar: Starting Your Net Business Without Venture Capital," published by Perseus in 2001. He has also written "Learning Economics," a collection of essays on economic issues. Kling 's personal web site is http://arnoldkling.com. His blog (with Bryan Caplan) is at http://econlog.econlib.org. He teaches high school on a volunteer basis near his home in Silver Spring, Maryland. He is married, with three daughters."

Jerry Martinson On How To Improve Biomedical Empricism in FuturePundit comments

I like what Jerry Martinson has to say here. Comments? Critiques?

http://www.futurepundit.com/mt/mt-altcomments.cgi?entry_id=4589

Jerry Martinson said at September 18, 2007 12:43 AM:
I've never really understood how public research is funded in any particular area. I don't know but I'm guessing it has a lot to do with the political connections of the principle investigators. To me, it seems that the result is a messy patchwork of studies that have limited statistical power to discover the unknown. When I read medical research papers, I'm kind of shocked at the culture of the experimental design. There's a great emphasis on a few "canned" types of experimental design that worship at the alter of Gauss. This culture emphasizes formality of a particular kind of investigation, at the expense of trying to discover the unexpected. It seems like there is little coordinated planning on how to record information in such a way as to make synthetic "meta-analysis" actually automated or convenient, much less useful and accurate. I worry that the culture of professors guiding underlings who need to show original research to complete academic requirements in a certain way is inhibiting genuine discovery.

I'm not aware of much philosophical work on the art experimental design for discovering the unknown that has influenced medical research. There is quite a bit of work in electrical engineering and in certain parts of OR-centric industrial engineering that have done quite a bit of work on the problem in general however. Consider the hidden Markov models or Kalman filtering extensions that electrical engineers have to learn in school. I remember in college where fresh after learning about Kalman filters, Markov models, and Bayesian methods in the EE classes, I then attended a very interesting lecture from George Box where it seemed like all these extremely powerful ideas that EE's are trained in could then be adapted to systematic empirical discovery of financial and medical models. The ideas seemed much more powerful than Black-Scholes or the simplified statistical methods that are taught in the life-sciences.

It would be interesting if one could develop a well-accepted method of making complex system models using this type of framework that mimic the complicated models we already know about in biological systems. We could then genericize the models and then we could then test various kinds of investigation and experimental designs to see how rapidly they converge on the correct hidden model. From the knowledge built up from that, places like the NIH or the HHMI could then make more intelligent funding decisions and set better standards for keeping and recording information.

Another thing about studies is that the published results create too much prose centered around discussion of the "significant" finding with the underlying valuable data archived in some arcane format that is inaccessible or incompatible with other works. Better standards for how the (unplublished) records are kept, organized, and made accessible could lead to much better derivative data mining efforts. Guidance from experimental design research could suggest which "schemas" would be best for derivative research rather than being a fool's errand for some SQL expert who knows little about the data he's organizing.

Should an expert engineer be fired for racism against a supervisor?

Excerpts from the article and my commentary below:

http://seattletimes.nwsource.com/html/boeingaerospace/2003889663_boeing180.html

"Weldon was fired in July 2006. He alleged in a whistle-blower complaint with the U.S. Occupational Safety and Health Administration (OSHA) that the firing was "retaliation for raising concerns throughout the last two years of his employment about the crashworthiness of the 787."

But according to a summary of OSHA's findings, Boeing told investigators Weldon was fired for threatening a supervisor, specifically for stating he wanted to hang the African-American executive "on a meat hook" and that he "wouldn't mind" seeing a noose around the executive's neck.

Weldon denied to OSHA investigators that he had referred to a noose and said the "meat hook" reference had not been a threat.

OSHA dismissed Weldon's claim, denying him whistle-blower status largely on the grounds that Boeing's 787 design does not violate any FAA regulations or standards."

My commentary: Let's set aside whether this is a frame up of a whistle blower. Weldon by all agreements seems to be a competent engineer well into career of strong positive contributions in areas we can all benefit. What rational sense does it make for him to be fired even if he did make racist comments to an african american executive? As a society, he's not valuable to us because he's good at promoting tolerance, he's valuable to us because he improves the state of our engineering solutions to problems we face. An advocate for african americans could point out african americans probably aren't getting disproportionate benefit from Weldon, so why should they bear disproportionate cost? The african american executive (his supervisor) could make a similar argument. But to solve the problem with the african american executive, Weldon or the executive could just be reassigned, with Weldon paying appropriate damages to both the executive and the boeing  for the costs both incurred as a result of his racism choice. In addition Weldon could in theory compensate the african american community as a whole, for example, by contributing a specific financial amount to a media and education fund which counters negative public representations of african americans.

But what do we gain from Weldon being fired, not because he know longer makes sufficiently good contributions as an engineer, but because he chooses to make racist public comments about african american individuals (allegedly)? It's a net loss, and as far as I can tell these type decisions only lower my persistence odds.

***

More broadly, this type of reasoning is why I favor allowing people to do "horrific" things if their contributions outweigh these horrific things. For example, let's say a very productive emergency room doctor can increase her productive saving of lives by 2 people every time she murders one of her patients randomly, up to 10 more saved people per month. Rationally, I think we as a society should allow, even encourage her to murder up to 5 of her patients randomly per month. We only benefit as a result.

I'd like to see the legal system and social norms reworked along this notion. I think this is sort of done when people suggest we should move from a criminal/punitive system to a torts/redistributive system. Feel free to explore this idea in the comments, and other general comments on this topic is welcome.

Is Mtravern preying on biases for status advantage in the overcomingbias space? (HA says yes)

http://www.overcomingbias.com/2007/09/human-evil-and-.html#comment-83064003

mtravern: "Hopefully Anon, I can't imagine what great status or survival advantages you think I'm going to get by posting here. If I was interested in advancing my status I'd pick better activities than flaming pseudonymously on blogs, you may be sure.

But in a way you are right, but trivially. Presumably everybody who posts here or otherwise engages in human communication has, at some level, an underlying motive of increasing their status. Everybody hopes to make themselves look good. I don't see how my postings differ from anybody else's in this regard.

Posted by: mtraven | September 17, 2007 at 03:00 PM"

Mtravern, I think it's objectionable the degree to which you're preying on bias for status in the overcomingbias space. It's a space putatively created to help us become better at modeling reality, and to be more effective at determining what actions our in our best interest. It's a corruption of that space to encourage ideas to win out by the degree to which they already conform with popular or powerful notions, or notions we may be genetically predisposed to support, rather than the degree to which they improve our ability to accurately model reality and make effective decisions based on those models.

It seems to me your internal calculus is: (1)to maximize my personal status, I want my name to be hitched to a winning idea in an interthread discourse. (2) In my every experience, opponents to torture win in discourse over proponents of something other than reflexive opposition to torture. (3) Therefore, I want to hitch my name to the anti-torture element of a discourse where some people support something than reflexive opposition to torture.

In real non-anonymous life, I like to free ride off the efforts of others as much as possible and so appreciate these type tactics for self-interest maximization. But spaces like overcomingbias are rare, they present the possibilities for new and better ideas and approaches to be innovated, and I think it's a disgusting tragedy of the commons (that likely increases my mortality odds) when free thought is stifled or bogged down by such craven attempts to "win" discussions by something other than really good ideas and models of reality -and I think your posturings in the discussion on torture are examples of such craven moves.

However, your response is welcome! (as are the thoughts of any 3rd parties reading this).

Engage me On Discussing Moral and Representational Hierarchies Critically

http://www.overcomingbias.com/2007/09/human-evil-and-.html#comment-82879465

mtraven,
You're ignoring concepts like inquiring empirically whether situationally employed torture can reduce net torture in the world. I brought it up in the torture thread and I think every commenter ignored the concept. Also, I'm unconvinced that the desire to label indivuduals as "good" and "evil" comes from a good faith attempt to accurately model reality, or even optimally solve existential challenges we face. It seems to come more from a desire to use morality to create status heirarchies, although in some cases it could also create representational heirarchies privileging both the people framed as "good" and as "evil", to the detriment of those that embody neither archetypes.

I'm surprised no one else in engaging me on these ideas of moral and representational heirarchies as ends in and of themselves.

Posted by: Hopefully Anonymous | September 15, 2007 at 12:20 PM

Response to Recen Anders Sandberg and TGGP Posts: Mao, Stalin -the most evil?

Some recent Anders Sandberg and TGGP posts:

http://www.overcomingbias.com/2007/09/human-evil-and-.html?cid=82802703#comment-82802703

In my post here you can find links to defenses of Stalin and Mao. They do not deny that both killed a huge amount of people, or that a great many of those people were completely innocent and it was tragic that they were killed. The author instead states that Stalin and Mao gave enough benefits to people, who had previously been faced with even worse rule, that those unfortunate deaths should be viewed as cons outweighed by the pros. I can't do an effective job of presenting his case (for one thing I completely disagree with it and think the two were among the worst people in history) but I would like if he could explain more here, including why our view of them is so off.

Posted by: TGGP | September 14, 2007 at 04:27 PM

While Eliezer and I may be approaching the topic differently, I think we have very much the same aim. My approach will however never produce anything worthy to go into anybody's quote file.

Posted by: Anders Sandberg | September 14, 2007 at 04:48 PM

Anders & TGGP,
Response on my blog within 10 minutes (so as not to flood overcomingbias with posts).

Posted by: Hopefully Anonymous | September 14, 2007 at 04:51 PM

My response:

Anders, in my opinion you come across as being politic. Probably a good strategy given that you and Eliezer are both non-anonymously posting at overcomingbias. I encourage you to have another, anonymous blog where you free yourself from politic responses and give your best criticisms possible of the ideas of others. Think about how it could move discourse to maximize our persistence odds.

TGGP, do you really think of Stalin and Mao as "among the worst people in history"? Because that would seem to me by almost any measure to be a suboptimal outcome for them. It places them at the near bottom of a moral hierarchy, and almost everyone who ever lived above them in that heirarchy. Beyond that, how sure are we that they even had free will (or that we do)? Moving from moral hierarchies to representational heirarchies, they still make out pretty good -after all, the devil had at least 2nd billing in medieval morality plays. But I think there are richer modes of analysis for us than "Stalin and Mao, very bad. Louis Pastuer, very good" -except to the Straussian degree it can positively motivate other humans around us to work to maximize our mutual  persistence odds.

I rail yet again against the concept of "both sides" rather than n-th possible sides

http://www.overcomingbias.com/2007/09/what-evidence-d.html#comment-82800843

Phil: "Isn't it reasonable to assume that we might be predisposed to being biased in favor of whichever argument we have heard most recently, and that we should withhold final judgment until we have heard both sides?"

The concept of "both sides" may be the mother of all biases. Primate social groups, in my understanding, often make leadership decisions by an alpha male and a challenger male battling for supremacy. This bias of decision-making by weighing 2 sides (as opposed to considering there may be n-th possible 'sides') warps a significant amount of decision making, it seems to me.

Posted by: Hopefully Anonymous | September 14, 2007 at 04:34 PM

Praising Anders Samberg, Criticizing Eliezer Yudokowsky, On OvercomingBias

http://www.overcomingbias.com/2007/09/human-evil-and-.html#comment-82798141

Eliezer, do you see the distance between how you are discussing this topic and how Anders is? Invoking "Good and Evil", with Stalin and Mao (losers of conflicts with the anglosphere?) being "evil" seems to me to me to be an appropriation of the overcomingbias space more so than a good faith effort to intelligently expand that space. Anders approach in general and in this post in particular serves as a good contrast, in my opinion.

Posted by: Hopefully Anonymous | September 14, 2007 at 04:09 PM