I think it’s possible that algorithms and computability are simply the wrong framework for understanding human intelligence or creating artificial intelligence.

They probably are the wrong framework. But I don’t see how that helps you with the halting problem. The issue with the halting problem isn’t that a solution is really hard, or that it’s ill-suited to the methods of computer programming. It’s that the notion of a correct general solution to the problem is, provably, internally contradictory. There’s nothing that could count as a correct general solution to the problem.

Here’s why. The halting problem is provably unsolvable because if it were solvable by any means, its output could be used to construct at least one algorithm which would be logically incompatible with the correctness of the solution to the halting problem. For a brief discussion why, see here. Basically, any proposed solution to the halting problem could be used to construct a “fink” function, which will do the reverse of what the halting problem says a given function will do when given itself as an input (so that, when you call FINK(FUNC), execution will halt if FUNC(FUNC) would loop forever, and will loop forever if FUNC(FUNC) would halt.) But then, when FINK(FUNC) gets itself as an input — FINK(FINK) — there is literally no possible answer to the question of what it will do. If it halts, then it doesn’t halt; if it doesn’t halt, then it halts. Hence, whatever your solution to the halting problem said it will do, it won’t do that. Hence, whatever your solution to the halting problem is, it’s wrong in at least one edge case.

Note that it does not matter whether the solution to the halting problem is accomplished within a Turing machine, or whether it is accomplished by some other means. Suppose the method for solving it is to print out a copy on a human being’s printer, and then wait for the human being to input TRUE or FALSE, and send the human being’s response back to the function to use as the return value for the halting function. Even so, you can still construct the fink function using that external input, which means that there just is no right answer that the human being could possibly return: whatever answer she returns, the program will do the reverse of what she said it would do.

The problem is not that finding a method to solve the halting problem is really hard, or limited by available resources or conventional computer architectures; it’s that the mere existence of a solution would necessarily entail an edge case where the solution cannot possibly be correct. A human being can’t do it any better than a computer can.

Yeah. One of the notable subcategories of failed predictions, which, in the interest of brevity, I didn’t really discuss at much length, are the number of things which Kurzweil predicts which actually are now feasible or already implemented, and which have in some sense broken through — but as permanent niche products, which remain popular with a select class of people for a select few applications, and which pose no foreseeable threat of displacing anything or taking over the world like Kurzweil imagined they would. Reliable speech-to-text software is another example — you can pick up Dragon NaturallySpeaking for a couple hundred dollars and it’s good enough that a number of professional writers use it to do more or less all their writing. It’s just that all of them happen to have carpal tunnel syndrome or severe motor disabilities. Just about everyone else still uses a keyboard, even on devices like iPhones that don’t even have a physical keyboard (but do have a physical microphone, which could have been used for speech-recognition if anyone cared enough to insist on it).

It’s not that these technologies aren’t available, or even that they are priced out of most people’s reach; it’s just that people have the technology now but most of us don’t have much of a use for it as of yet. And there are, of course, drawbacks and trade-offs involved (you may love the idea of dictating your latest memo; but do you really want to hear all of your coworkers dictating theirs at the same time?), which futurists characteristically fail to take any account of.

Hi,

The problem that you ran into is what I’ve come to call the “upgrade downgrade” problem: FeedWordPress strongly encourages you to upgrade MagpieRSS to a newer and much more capable version than the one packaged with WordPress, but when you upgrade your version of WordPress, one of the files that you get in the upgrade package is a brand new copy of the old and busted MagpieRSS, which overwrites your upgraded version. In the most recent versions of FeedWordPress. I discuss the problem (from back when WordPress 2.5 was released) here: http://feedwordpress.radgeek.com/blogs/radgeek/2008/04/18/upgrade-downgrade. Current versions of FeedWordPress should put up a loud warning in the WordPress Dashboard if you log in after inadvertently reverting your copy of MagpieRSS. Let me know if this failed to happen for you; if so, I’ll investigate a bit to try and figure out why you weren’t notified of the problem.

Eventually (as in, within about 2-3 releases from now), I intend to transition FeedWordPress from MagpieRSS to the much more advanced SimplePie package which is now included with newer versions of WordPress. But it’s going to be somewhat tricky because the existing filtering interface is based around MagpieRSS implementations, so I need to write a compatibility layer if I want to avoid breaking other people’s filters.

If you’re curious, the reason you got posts with just capital “A,” in particular, it was probably because the feed you were syndicating included MediaRSS extensions, and in particular a media:content element. (WordPress.com feeds, for example, include MediaRSS meta-data by default.) The old version of MagpieRSS had a bug in its parsing code that mangled post contents if it encountered any other element named “content,” even if it was in a different namespace. My customized branch of MagpieRSS eliminated the bug, but, as mentioned, when you upgraded WordPress it temporarily reverted MagpieRSS back to the old and busted version, which lost that upgrade.

goffchile:

I would argue that there *are* significant entry barriers–owning a computer, internet access, and time being huge

I think the latter is by far the most important. In more or less every city, the availability of Internet terminals in public schools and public libraries means that not owning your own PC, or not being able to afford home Internet access, is no longer a significant barrier to web-based applications like blogging. Working 60 hours a week at three different jobs, on the other hand, is.

Besides the barriers on the supply-side, the other important concern (which a lot of feminist bloggers, for example, have raised) is on the demand-side. As many millions of blogs as there may be, attention in blogging is structured much more hierarchically than blog boosters are inclined to acknowledge, and that hierarchical structure much more closely reflects traditional social hierarchy than they care to admit (actually, often, a hyperthyroidic version of traditional social hierarchy, because straight white male educated professional “A-list” bloggers have, so far, been subjected to critical scrutiny far less than straight white male educated professional “MSM” outfits).

Unfortunately being able to speak is fairly irrelevant, from the standpoint of politics or civil society, if nobody hears what you have to say, or nobody takes it seriously enough to consider it worth listening to. I think that blogs are a move in the right direction — and one which will become increasingly important with time — but there’s a long walk down and a long, hard slog ahead between that mountaintop to the Promised Land.

I should say that in the medium to long term, actually, I think that what will be far more important than any blogger’s ability to show up on the mainstream media’s radar, or even to break through into “A-list” bloggers’ boys’ club mutual linking society, is that blogs are making it much easier for writers with a distinctive view to simply bypass broadcasting prominence and to reach a smaller, mostly self-selected audience with more narrowly focused interests. As people change their habits of reading, conversing, and news-gathering, broadcast success will become less and less relevant, and deadlocked mainstream consensuses will be shifted because, by nearly imperceptible steps, the ground collapses out from under them, not because some mighty force erupts up through them. But, again, we’re still a long way from that, and I think that part of the process of getting to that will involve recognizing how far we are from it and consciously changing our tools and our habits (in both reading and writing) to work towards traversing the gap.

Start using public key cryptography to the extent possible in your private correspondence. Encourage others to do so. Help non-technical users get started with it.

Support your neighborhood CopWatch. If you don’t have a neighborhood CopWatch, get in touch with the nearest one and ask for their advice and/or help in starting one.

Write to public forums that don’t usually publish anarchist material, but where you stand a chance of getting published anyway, explicitly advocating anarchism. For example, a letter to the editor of your local newspaper. Include pointers to online resources where people can learn more about anarchist takes on the issue you’re writing about.

Find ways to get things that you need outside of the documented cash economy. For example, if there’s a Food Not Bombs in your town, you can get to know a bunch of other anarchists, do some mutual aid work, and, in the process, get some free food for your labor. If there isn’t a Food Not Bombs in your town yet, again, get in touch with existing FNB groups and try to start one. (There are lots of guides online.)

If you have it, I’d also recommend contributing some money to groups that provide direct relief and aid for victims of violence, and which combat cultural attitudes that promote violence. For example, I give a fixed percentage of my income to women’s shelters and groups like Women for Women International (which focuses on relief for women in war zones). By doing so I not only provide direct aid to real people and to a network of institutions which can supplant the supposed welfare functions of the State; I also remove that much more money from the taxed economy, and put it toward the purpose of healing and mutual aid, rather than what it would otherwise have been used for — graft, handcuffs, bombs, prisons, etc.

This is a software bug in feed consumers, related to some mistakes, rather easily made under certain conditions, as to how to handle Yahoo’s MediaRSS namespaced elements. It is not, essentially, a problem with WordPress.com’s feeds.

FeedWordPress users may be pleased to know that this bug has been fixed in the latest release of FeedWordPress, version 0.991.