He lauded a FasTrak-McDonald's alliance, where drive-through McDonald's customers with FasTrak will be automatically billed for their purchase, saving a precious few moments by avoiding that extra transaction. I imagined hacking the system and charging everyone with FasTrak who came by my house $20 ... I wouldn't, but I bet someone could and would. What a privacy and security risk! (Furthermore, I'm a fan of slow food, and the concept of saving a few dozen seconds at a drive-through just seems ... distasteful.)
Another system that raised my privacy warning flags was a toilet that analyzed your "output" and sent the information wirelessly to your doctor. If it's in your doctor's office and wired into their network, fine ... but Jean-Paul suggested that they could be everywhere. This reminds me of some of the issues with supermarket cards - HMOs in Great Britain are already moving to use buying information to deny health insurance, and American HMOs may follow suit. (More info about supermarket cards here, here.) I don't think I'd want my doctor and whoever else cares to intercept the information to know everything about my so-called output.
(A topic related to both of the above examples is privacy concerns with embedded RFID tags. Sure, it makes inventory so much easier - just point a sensor at a warehouse or store, and get counts of all the items inside - but I don't like the thought of being able to drive by my house and get an inventory of the products I have ... or even the thought of companies throwing directed advertising at me, a la Minority Report [even without the retinal scanning].)
How about using your cell phone to buy stuff from vending machines? How about vending machines that only accept payments from a cell phone? Not only do you have to have a line of credit, you have to own a cell phone to buy something from these machines. What percentage of the American population can afford to have both a line of credit and a cell phone? How about internationally? This is what I call the "digital maid" phenomenon - technology applied to simulate a digital maid, for those few who can afford it. Sadly, the vast majority of cool cutting-edge apps I see at Berkeley, conferences, and elsewhere are designed for this affluent (in global terms) market, because the designers themselves are in that market. All they have to do is design for their own needs, or needs of people that are very similar.
And much more ...
On a somewhat related note, in about 45 minutes I'll go to the first Engineers without Borders meeting for this semester. It'd be great to do a EWF internship next summer - last summer they had two internships for human-computer interaction students in Bosnia, and probably plenty that I didn't hear of.