Tic-Tac-Tome is a 1400 page policy for playing tic-tac-toe. Like a giant Choose-Your-Own-Adventure, the reader chooses a location to move to, and turns to the appropriate page to see the counter move. Of course, the book plays optimally, and so “the only winning move is not to play.”

The book fits perfectly the Chinese Room argument. In the thought experiment, a Chinese speaker writes messages in chinese and slips them under the door to a locked room. Responses, also written in Chinese get passed back under the door. The responses are so convincing, that the Chinese speaker is convinced he/she is conversing with an intelligence that understands Chinese. Unbeknownst to those outside, a person that does not speak Chinese collects the papers as they slide under the door, consults a giant lookup table of inputs to outputs and then copies the prescribed response to another piece of paper and slides it back, never understanding the inputs or the outputs. The question is then, whether Chinese speaker is conversing with an intelligence or nor, and if so where does the intelligence lie?

Personally, I find the whole “Is it live, or is it Memorex?” argument rather quaint and tiresome. Something that’s only worth discussing while riding in my atomic powered self-driving car while smoking a bowl of the finest hashish. AI always struck me a bit like a magic trick. From the outside, it looks amazing (Wow! You made an orange float in the air! Amazing!), then you find out how it is actually done, and then you’re disappointed because your fantasy has been dashed (You just shoved your thumb in it! You suck!). Personally, I think this says more about us, and our willingness to be misled than anything else.

Previously.

I’ll let you find the hallal video on your own.

It is the year 2011, and I now have a robot maid. Alas, it does not have a sassy voice, and even though it comes in sexy black, it does not come with a french accent either. Mostly it just beeps, like robots from a long time ago, but occasionally it does speak in a feminine voice, but only to chide me. (“Error one. Please move Roomba to a new location and press the ‘clean’ button.”) It’s definitely not fast, and it’s not thorough, since there are some places too small for it to fit, but I do like that I don’t have to vacuum. My favorite thing it does is how it desperately tries to claw its way back to the charging station if you try to drag it away so you can work on it. That said, I still will be eagerly awaiting the commercial arrival of a Mahru-Z like bot.

Opening the box, I found this sticker on one of the plastic bags:

This robot contains an electronic and software interface that allows you to control or modify its behavior, and remotely monitor its sensors. For software programmers interested in giving Roomba new functionality, we encourage you to do so.

Because the functionality of iRobot Roomba can be changed by you or other third parties, usage of this Roomba is subject to the enclosed End User License Agreement. If you do not accept this agreement, please do not open this package. For more information, visit www.irobot.com

Now that’s nice. More companies should take a pro-hacker stance. A “you assumed the risk” stance towards added functionality is great. If Igive them money, and don’t ask for support, then why should they care what I do? I’d love to be able to download a a more efficient floor covering algorithm to it, perhaps even have actually remember the shape of the rooms from run to run. Unfortunately, the hacking sites are filled with lameness like adding Knight Rider lights, and Maybe there’s a community of people doing useful things, but somehow I doubt it.

According to Lt. Col. Dave Thompson, USMC commander of cybernetic warriors, or chief robot wrangler, there are over 2000 military bots in Afghanistan. Doing the math, this means about one out of fifty soldiers are robots. It needs to be said, that this number is only the number of bots deployed, not the number that are actually utilized.

Also, Thompson was only counting land bots, not the aerial drones that get all the headlines, and movies. These bots are the bomb defusing Talons, PackBots, and few other mine clearing bots. Notably missing from Afghanistan are SWORDS, Talons fitted with M-60s and other goodies. Those were withdrawn from the battlefield after they went berzerk during their deployment in Iraq. (“Kill-bots want peace too; but programs must complete.”) Also missing from this list are the more advanced bots, such as self driving cars, and of course everyone’s favorite BigDog.

Local killbot proprietors, Survival Research Labs is having a show! It’s this Saturday (July 24, 2010) in Petaluma, at the Petaluma River Heritage Center (260 Water St, Petaluma). Admission is free, and is part of the Friends of the Petaluma River‘s Rivertown Revival.

It is going to be fun on the bun.

MIT has published a paper entitled Programmable Matter by Folding (full article) that describes paper that can fold itself into a variety of shapes. The paper is covered by is divided into triangular sections that are joined by a network of thin nitinol actuators that contract under voltage. At the center of each section is a magnet that is used to retain the paper’s shape.

While I’m sure MIT had bigger plans for this tech (Well if it was the Media Lab, perhaps not.), I immediately thought that this was the perfect thing for synthetic plants. I’ve been thinking about how nitinol wires, or at least something like them, could deform a paper but thought that the being able to compresses only about 4% was a problem. When I first saw this video, I thought they were using something else besides nitinol, but they’re not. The trick they they used to get 180 degree bending is folding and annealing the 100μm foil so that the nitinol will remember the folded shape. Once it cooled, the foil is manually flattened, and then reheating the foil with electrical current will cause it return to the folded shape.

Guess it’s time to get some nitinol sheets.

Now this is nifty.

We Make Money, Not Art interviews, Gilberto Esparza about his Plantas Nómadas (Nomadic Plants), an autonomous walking robot that is powered by a combination of solar cells and a microbial fuel cell. When the fuel cell output drops beneath some threshold, the bot seeks out a water source, extends a proboscis and refills the fuel cell. Additional water is used nourish a colony of on board plants.

Gilberto’s earlier work is equally interesting. Parasitos Urbanos (Urban Parasites) (flashless site) was a series of robots inductively powered from electrical transmission lines that would move through the urban environment mimicking sounds they encounter.

Previously.

Jim Blackhurst’s SmartLED SolarTherm is a minimalist information display. Consisting of an RGB LED, a watch, and an ATTiny25 microcontroller. The chip contains a temperature sensor whose reading is displayed as light pulses. According the comments on Makezine, the internal temperature sensor is +/- 10 C (+/- 18 F), so its not very useful.

SolarTherm is simpar to M27′s Zach DeBord’s pummers. These charge a capacitor from a solar cell, and when the light level drops, the capacitor discharges, and causes an LED to blink.

While as an ambient displays these are visually interesting, especially Zach DeBord’s pummers, these seem to suffer from the main problem with all ambient displays. They trade simplicity for usefulness.

I want the display to be both pretty, but also informative. The display needs to be immediately interrogated. Similar to the how a grandfather clock provides a chime ever 15 minutes to an hour, but also can be viewed in order to learn the exact time. I’m thinking of something like Riedi and Gloor’s Weather Diorama.

Things like Nabaztag or the infinitely more endearing, Michael Kaminsky and Paul Dourish‘s SWEETPEA (aka “The Microsoft Barney Paper”) are more confusing than anything. Even baseball signs aren’t that confusing.

Maybe the best ambient display I’ve seen was simply a string hanging from a DC small motor wired directly into an ethernet cable. As packets would pass, the motor would be powered, causing the string to wiggle. As the network activity increased, so would the vigorousness of the string’s dancing. The great thing about this display is that it’s immediately and intuitively interpretable, while something more complex requires the user to learn some of sign language.