Saturday, December 17, 2005

...the work week will be 24 hours long.

Over the last 100 years technology has enabled people to get more done in less time. At home, we have household appliances such as dish washers, washing machines, and microwaves. At work we have computers, e-mail, video conferencing, and high speed internet connections. Don't get me wrong, we don't always like these devices. Windows crashes, our inboxes fill up with spam, and everybody's PC seems to have joined the virus of the month club. Even as I write this, the spin cycle on my out-of-balance washing machine is registering 3.2 on the Richter scale.

As frustrating as technology can be, it's hard to argue that we aren't more productive because of it. So, if we can get more done in less time, why aren't we working less? In the 50's and 60's futurists promised us lives of leisure. They thought our every need would be tended to by robots while we lounged in a hammock sipping colorful drinks shaded by a small umbrella.

There were two large problems with this idea. The first being, it's 2005 and certain technologies haven't advanced as quickly as expected. Artificial Intelligence has proven to be a very difficult problem to solve (the technology required for a computer or robot to understand what it's being asked to do and respond appropriately). Image and voice recognition has been equally frustrating. Although reliable systems have emerged for certain specialized applications, there's not a computer out there that can recognize objects and sounds the way a human being can.

The other large problem with the idea that our work will be done by machines is this - there is no arbitrary measure for how much work a person should accomplish in a day. We live in a competitive, capitalistic society. You're measured against the people standing next to you. If they get more done, you have to compete. As such, we now work just as many hours per week, if not more, as we ever have. The times may eventually change though.

In 2002 a robot named Grace completed a unique challenge. It was dropped off in front of a conference center full of people. Using voice and image recognition, it made it's way to a registration table by asking directions rather than being guided by an operator, After registering and then finding its way to an auditorium, it gave a short speech about itself.

The technology is still young but ask yourself - how far is this robot from being able to ask if you'd like fries with your burger? If packaged in a body capable of moving like Honda's Asimo, how far would it be from doing many jobs requiring unskilled labor? This is great right? We'll all be able to have our own personal electronic servants. They'll be able to act as a butler, maid, personal shopper...

Well, not exactly... or maybe eventually...

I think it's safe to say that these machines will be extraordinarily expensive when they first hit the market. They'll be used in very specific niches such as search and rescue or health care where their expense can be justified. Eventually, the cost will start to fall though. Let's say the cost falls to 40,000 of today's dollars. Who will have them in their homes? A handful of wealthy individuals and fanatic early-adopters - not many people overall.

At $40,000, who else may buy them? Well, what does a typical fast food worker make these days? I know there are a handful of places around here that start at $10 per hour but I'm guessing $8 is closer to the norm with places hiring as low as minimum wage. We'll use $8 as an average. Let's assume that most places are open approximately 16 hours per day. For one employee to be working during that entire 16 hours, it costs $128 in wages alone, not including social security matching, benefits, unemployment, etc. That's well over $40,000 per year. Even if that one employee could be replaced by a $100,000 robot, the cost savings would be worth it in a couple of years.

What we are likely to be facing instead of a life of leisure is an unemployment crisis. We've seen something similar happen in the manufacturing industry with a combination of automation and outsourcing to eastern Asia and Mexico. We may be on the verge of another similar event in the IT industry by outsourcing to India. This, however, will likely be far beyond the scope of anything we've seen before.

I expect first it will happen in an industry such as the fast food industry. We'll find out that McDonald's has outfitted a handful of stores with prototypes to try the idea out. People may balk at first and complain that they aren't dealing with a real person - especially if the technology is rolled out too early and buggy. Eventually though, the kinks will be worked out and people will find that there are some advantages. The machines will never cop an attitude. They'll never be off on a smoke break when you are ready to order. They won't give you back incorrect change or mess up your order. It's an even better deal for McDonald's. They'll be cheaper. They'll never steal. Once they're eventually given cooking duties, they'll never undercook a burger and make anybody ill.

Once these machines have been successfully deployed for one line of work, people will soon start adapting them for other types of work. In fact, I expect that our definition of skilled vs. unskilled labor will get a major overhaul in a relatively short period of time. Perhaps we'll find that painting is easy for a machine to do but framing a house is difficult. Perhaps we'll find that roofing is difficult for a machine but excavating is a piece of cake. We're almost certainly going to find that a machine can't replace a gourmet chef but it may be able to replace a short order cook. It probably can't sell a car but I'd bet it could wander the isles of Wal Mart helping people find things.

Regardless of which jobs become automated, it's almost certainly going to be a ton of them. If the transition is slow enough, maybe we'll see most people adapt smoothly into other lines of work. If it is rapid though we may see boycotts, mass protests, and riots by a large body of displaced workers. How do we allow progress to continue while protecting people's livelyhoods? Better yet, how do we capture a slice of the future we were promised by 20th century futurists.

I'd say one of the most promising yet difficult solutions to implement would be cutting the work week back to 3 or 4 days per week with the same pay as 5 days. The remaining jobs would be spread between more people reducing unemployment and everybody would have more leisure time which would probably be a boon to recreation based industries.

The only problem with cutting back the work week is actually implementing it. Companies have an incentive to squeeze as much out of their salaried employees as possible to compete with other companies. Even if they were forced to cut work hours through legislation, individuals would almost certainly take on two jobs for the additional income. In the end, even though most people would say they'd love to have 3 or 4 days off a week, if you give it to them, they probably won't actually take it.

What to do? Well, we've probably got at least 20 years to figure it out, maybe more. I say we start learning how to work less, appreciate the extra productivity that our tools have created for us, and enjoy some extra time with our friends and family.

Monday, December 12, 2005

...everything will be wireless.

including electricity

Sunday, December 11, 2005 wife will be able to work the TV

When TV's were invented, they were created to do one thing - capture audio and video signals that were being broadcast over the airwaves and turn them into a moving picture accompanied by sound. There were only two things that you ever hooked up to a television: electricity and an antenna. Things changed quickly though in the 70's and 80's. Along came video game systems, VCR's, cable TV, home computers. This introduced a small problem. How do you hook all of this stuff up to one TV? Assuming you had the right combination such as cable, a VCR, and a Nintendo, you were safe. You could run the cable through the VCR then run the VCR through the Nintendo RF converter. The VCR could record the cable at all times and if you turned on the Nintendo, it would always take over the screen.

Well, things have gotten much worse over time. In addition to the devices mentioned above, we now have TiVo's, camcorders, DVD players, cable boxes, HDTV tuners. RF converters are still around but we also have composite video, S-video, component video, DVI, HDMI. To make matters even worse yet, the popularity of home theatre and surround sound systems has broken the audio out of most of these video formats. We have analog audio, coaxial digital, s/pdif, optical digital, and a number of oddball proprietary formats. How do you hook all of this stuff up to your TV? If you have a separate sound sytem, how do you also hook it all up to the sound system? It's a bloody mess.

My HDTV has 8 inputs and I'm using most of them.

1) Antenna - I don't actually use this but it picks up over-the-air broadcasts.
2) Video 1 (Coax) - TiVo
3) Video 2 (composite video + analog audio on the front panel) - occasionally used for camcorder or other portable device
4) Video 3 (composite video + analog audio + s-video) - VCR
5) Video 4 (composite video + analog audio + s-video) - Macintosh mini
6) Video 5 (component video + analog audio) - HDTV Cablebox
7) Video 6 (component video + analog audio) - XBox
8) Video 7 (HDMI) - upconverting DVD player

This is a fair setup but it has it's problems. For example, the Mac mini should be on a DVI to HDMI converter and I should be viewing it in high resolution. As-is, the text is hard to read and I have limited screen real estate. Also, I have old game systems that I'd love to have hooked up for nostalgia. There just aren't enough ports level on my TV. Technically I could buy a switch box but AV switchboxes that are capable of converting between the multitude of formats are very expensive - especially if they support high definition. Plus, I'd be adding another level of complexity to an already complex system.

So far, we've only talked about my video setup. I also have a surround sound system. It has two optical inputs, two analog inputs, and one digital coax input. That doesn't match my TV so what do I do? Luckily my TV has an analog audio out. Unfortunately, the sound quality from it is attrocious. It produces a very loud background hum. Essentially, I'm forced to only use my surround sound system when I'm watching a DVD (using one optical input) or watching cable (using the other optical intput). I also have the TV's analog out run into one of the analog inputs but I rarely use it because of the annoying hum.

There are two other minor problems to be considered here. HDMI carries audio and video in one cable. How to you split it off to separate sources? Of course, any device that has an HDMI output is likely to have other outputs as well but it's no guarantee. And what is the cost of having so many inputs and outputs of various types on my devices? It certainly raises the price of these things.

So, it's a bit ugly but I've waded my way through these rapids and gotten everything more or less connected to my TV and my sound system. It may have been a bit confusing and there may be a lot of wires running around behind my entertainment center but all is good otherwise right?

Operating all of these devices requires 6 remote controls.

My wife, having not hooked all of this together is hopelessly lost. She can switch between the cable and DVD player but that's about it and if anything is left at all outside the normal configuration, there's no guarantee she'll even be able to do that much. In some cases, she may need to change settings on multiple devices in a coordinated fashion to get the desired result (for example, switching video source on the TV, switching audio source on the sound system and setting audio output settings on the desired device). I believe we have the technology to fix this mess if the big players are willing to come together and work on it. Unfortunately, I think we need yet another audio/video format to do it.

Problem #1
You don't know how many devices people will be hooking up to their TV.

They may have just a cable box or they may have 20 different electronic devices that they want to use. I suggest this should be solved in the same way that it was solved for firewire. Allow devices to be daisy chained together. Not only can I hook an indefinite number of devices together but I can do it with less cord. Devices that are stacked on top of each other can be connected with short 6" to 12" cables.

Problem #2
Selecting signal sources has to be coordinated across multiple devices.

This may be a little tricky but it should be perfectly doable. Every device should be registered as a signal source and/or a signal reciever. Recievers can be set up in groups so that when a signal source is selected they act in a coordinated way and all switch to that signal source. In addition, signal sources could provide a lower resolution preview video so that rich graphical menus could be created by TV manufacturers.

Problem #3
Too many remotes

This is by far the toughest problem here to solve because each device may have specialized functions. I think the ideal solution is probably to put those functions onscreen. Create one remote with nothing but volume, channel, cursor arrows, and enter/cancel buttons. Allow the devices to communicate with each other over the daisy chained data cables to share menu text and user selections.

If some sort of a system like this isn't adopted, things are only going to become more and more confusing as we move into an era with even more home audio and video devices. And if you don't believe we'll have more, my new speaker system actually hooks into the TV to provide configuration menus. Who'd have thought?

...I'll never lose my car keys

In the near future, all of the things we buy will contain an RFID tag (Radio Frequency Identification). These are tiny little devices that transmit a radio signal capable of traveling a short distance. Like a barcode, an RFID tag identifies a product. Stores will use them at checkout lines instead of scanning a UPC symbol. Unlike barcodes, it doesn't just identify the type of object (30GB iPod Video $299). It identifies the exact object (30GB iPod Video serial number 92384776 $299). This should also make them useful for inventory and anti-theft purposes. It also could have some negative impact on personal privacy but that's another topic of conversation.

So, once this technology hits the shelves, why can't I have my own personal RFID scanner that I can use to locate things around my house? When I buy something that I might accidentally lose, I scan it with my little RFID scanner and it adds it to an internal database. When I need to locate that something, I tell my scanner what I need to locate and walk around the house until it beeps or otherwise signals that I'm close.