I take this as another sign that the first true machine intelligence will emerge in the Googleplex.

The Google Box mesh network can reach out to nearby neighbors, too, bringing them onto the Internet in a way that would be difficult to stop or control even if the broadband ISPs wanted to, which they won’t, because Google will find a way to share the wealth with them. It is not in Google’s interest to put out of business any ISPs, so they’ll try hard not to. But it IS in Google’s interest for there to be universal broadband coverage, which the Google Cubes will, for the most part, enable.

When Google announced their intention to submit a proposal to provide wireless service within San Francisco, I thought they might go this route. Alas, their proposal turned out to be a little more conventional. But from a little blurb in Macworld magazine, I learned of the FON Wi-Fi service, which is building a Wi-Fi network based on this crazy, potentially revolutionary approach.

The idea is fairly simple. Many people, especially in dense urban settings, have broadband connections to the Internet, which are not fully utilized. If you were able to attach a smart wireless access point to each of these connections, you could provide coverage for your service, without investing in a lot of infrastructure. The wired part is already in people’s houses, and represents the majority of the physical infrastructure required; all you need to do is get lots of people to buy and install your smart box. You do this two ways: discount the box, and offer people an incentive to keep them running.

FON’s service is basically just this simple. There are two ways to get a FON box installed in your house. First, you can download software for your existing Wi-Fi router (Linksys and Buffalo products, mostly), and convert it to a FON box, at zero cost to everyone. Second, you can order a cheap Wi-Fi box from FON with the software pre-installed. Just unpack and plug it in.

The incentives to run a FON box are equally simple. First, if you opt to make your box open to other FON members for no compensation to yourself, you get free Wi-Fi access across the entire FON network. Anywhere you can find a FON Wi-Fi signal, you can connect to the Internet, for free. There are more than 80,000 “FONeros” worldwide, so this is a pretty strong offering already.

Alternatively, you can be compensated when people use your FON box to connect. (A day pass costs $3 to non-FONeros.) You split the revenue generated by your FON box 50/50 with FON. If you’re in a good location, that could add up to a nice chunk of change, certainly more than the FON box cost to acquire and run. If you run a coffee shop, or live next to a coffee shop, it seems like a no-brainer. Especially if you already have Internet access.

The secret sauce that makes this service work is the software inside the FON box, and the central authorization and billing service they connect back to. (The magic, and potentially huge profits, comes from Metcalfe’s Law.) It’s too early to say if FON will be a resounding success, but I’ve certainly ordered my FON box.

Two other interesting points. First, like Google in the Cringely article, FON is planning to make their own boxes. The first will just be a simpler, cheaper version of the Linksys box most people are using today, but the second will include an integrated Skype phone handset. If that doesn’t make your head spin…

Second, FON’s investors include Google. Imagine that.

An interesting, creative, and very persuasive use of the tool that Google created, and an illustration of how ease-of-use can change the game. Looking at satellite imagery before was either hard, expensive, or both. No longer. Anyone can do it, easily and for free, entirely on a whim. I predict similar “essays” will have a significant impact on political conversations around the US in the next two years.

Today I found a tool by Paul Rademacher that takes it one better, and showed how data from craigslist.org could be combined with the Google Maps API to create a very useful, interactive tool for shopping for rental apartments and real estate.

This is quite a step beyond using now-easily-accessible map pictures to illustrate a point; it combines data from two unrelated sources, and uses the power of Google Maps APIs to build an entirely new tool, something that Google may not have envisioned when they created Google Maps. No deals, no cross-licensing, no bizdev guys schmoozing endlessly — just one guy with an idea quickly hacking out some clever code to create something genuinely useful. This is how Google is attacking Microsoft, and in the long run it will probably work.

Of course, the power of Google Maps can be taken too far. I’m a little concerned about “abuses like this”:http://www.defectiveyeti.com/archives/001238.html.

Update: I’m no longer sending out invites, not necessary, you should be able to sign up with Gmail directly.

First of all, Google is very up front and direct about how Gmail works, what information they will have about you, and how they plan to use it. If you do not like the way Gmail works, you don’t have to sign up for it. Gmail is entirely opt-in (indeed, it’s currently hard to get the opportunity to opt in).

Why a state senator feels the need to “protect” people from something they don’t have to sign up for is beyond me. More proof that politicians are publicity hounds first, uninformed technophobes second, and advocates for genuine public good dead last.

Some of us, dammit, want Gmail to do what it does: scan your e-mail to improve the quality and relevancy of ads served. Log on to Yahoo!‘s or Hotmail’s web mail systems, and tell me you actually like the horrible Atkins and dating services ads they are serving up. Tell me that’s better than targeted ads that might actually be useful.

(Before you scoff about ads that are useful, read why I like Google’s ads. I’ve bought things from the ads placed on Google searches.)

The other response to these so-called privacy advocates is more technical. The concern is that Gmail will be scanning all incoming and outgoing e-mail. Well, so do all of the other web-based e-mail services. For that matter, so do most ISPs and corporations. They are scanning for viruses, worms, and spam, but they are scanning all your e-mail already.

As usual, Tim O’Reilly has the most sane and forward-looking take on the whole issue. For my part, I’m happy to make the deal with Google, relevant ads for a great web mail service (and it is noticeably better than Yahoo!‘s e-mail service, which I’ve used extensively for many years). My only real concern is, how do I get my archived e-mail (all 250,000 messages) into it?

But two articles over the last two weeks made me think of something different, namely, that the links that are spontaneously happening between the many active websites (mostly blogs) are similar to the formation of “neuronal paths.” We are building a giant, global brain.

The first article was Ben Hammersley’s Trackback in the Saddle Again, which describes quite a number of different ways that articles can semi-automatically form linkages to each other. The second was an interview with Cameron Marlow, the creator of Blogdex, and in particular his Social Network Explorer. Both articles illustrate how complex structures are being formed in the nodes (sites) and paths (links) between nodes.

So my thought today was, blogs and news sites are the neurons. Links from one to another are the neuronal paths, or connections between neurons. (TrackBack and its ilk are especially cool here, because the connections are two-way.) And services like Blogdex, Daypop, Social Network Explorer, and especially Google, are how the paths are reinforced to create “preferred” pathways through the internet, which alter future queries and explorations by making those paths more likely to be reused.

What’s especially neat about some of the services like Daypop is that the pathways are reinforced not only by people publishing on the web, but also by people who are just browsing the web. That is, when you follow a link off the Daypop Top 40, it records it, and that will influence how popular that link is, and therefore how many other people will see it, how long it will be around, etc.

These structures are certainly influencing how people think, if only because news stories are now frequently emerging in the blogsphere first, and once they’ve become sufficiently interesting and widespread, they cross over to the mainstream media. One could make a case that that’s how Trent Lott fell from grace.

Powerful stuff, but it’s a long way from having a giant brain. But that’s OK, on the scale of things, websites vs. neurons, there’s not enough nodes on the internet yet. Google indexes about 3 billion web pages today, but most of those don’t participate in the kinds of services described above. Blogdex and Daypop each index less than 20,000 blogs or news sites. So we’re talking pretty small numbers, today. There are thought to be 100 billion neurons in a human brain.

So our “giant” brain probably won’t be waking up and taking over the world tomorrow.

So I tried what I usually try in cases like this: I searched for the error message in Google, and read through the postings and solutions on message boards and archived mailing lists until I found something that worked for me. Unfortunately, this not uncommon error message was reported via a web page on my server — as it would be by any web server using PHP to access MySQL which had a problem. Can you see where this is going?

Yes, indeed, there are thousands of web sites with fucked up configurations spewing the same damn error message. All being spidered by Google. And being returned in my results. Out of 500 or so search hits I reviewed for the error message (out of 7310 hits total), only two were actually not an error page, and of those, one was in Russian.

Grrr.

The fact that you can read this means that I solved the issue, but it took me the better part of two days to solve, and in the end, the only way to fix things I could figure out was to re-install MySQL from an official binary distribution, not my preferred way to install software on my FreeBSD server.