Why Hal Will Never Exist

aengblom writes "Researchers at the University of Maryland's Human-Computer Interaction Lab are suggesting what many of us have already guessed. The future of human-computer interaction won't be through speech--it will remain visual (they explain why). The Washington Post is running a story about the researchers and how they think we will get computers to do what we want. The article is a fascinating read and is joined by a great video clip (real or quicktime) of the researchers and their methods. The Post is holding an online discussion with the researchers tomorrow. Also check-out Photomesa the lab's software program that helps track images on a computer. (Throw a directory with a 1,000 high-res files at this thing and you can justify that pricey new computer you bought)."

Comments

Photomesa.

[MisterBlister]

Great this will work well with all my porn.

Also Linux users are gay.

Hmmm. Photomesa...

[Chasing Amy]

Photomesa looks interesting. I could certainly use a better image browser, too--my pr0n directories now have 2,143,629 files, and counting... Sadly, I'm not kidding! Neatly categorized by sexual acts, scanners, or physical features, too. :-)

Currently the best image browser I have is ACDSee, a Windows app. Are there any better ones out there, for either Linux or Win32? Since the Photomesa image browser was mentioned in the parent article, this should qualify as an on-topic question. :-)

As for interacting with computers--I've long said that until AI is sufficient for computers to understand basic human speech at a fairly high level (which clearly is several decades away, at the least), that there are few improvements that can be made on the fundamental visual metaphors we already use. Many people are keen on 3D workspaces and such, but let's face it--they have no fundamental advantages over current 2D workspaces, except for the exceptional case of 3D modeling and similar spatial apps. But for browsing a filesystem, for example, I can't see any advantage that 3 dimensions have over 2, and it adds complexity and unneeded possibility for confusion.

There's a reason we've pretty much been using the 2D WIMP desktop fr over 20 years with few fundamental changes--it's a natural idea, more like a "discovery" than an invention. While there are many ways to implement the WIMP system, all have fundamental similarities and I doubt they can be bettered by another paradigm, such as 3D interfaces. Sit back and visualize different possibilities--I have, a lot, and I really can't see a paradigm that works better, except of course for the CLI for those who are willing to internalize computer functions in their memories rather than let the computer externalize them into visual interface.

That's just my opinion, of course. Oh well--back to organizing my pr0n. ;-)

Re:Things To Do Today

[Things To Do Tuesday]

The Chinese would have you practice otherwise. Puhsonally, me rike dry around with trophy blonde ballast gull.

Now, I have things to do Tuesday.

1. In an Integra or Civic, switch four lanes
2. With the pedal down, peel out, racing ain't a thang

Categorizing Images...

[Chasing Amy]

Most of my files are categorized fairly well because I've done a fairly good job of keeping up with new files at the end of each day I've downloaded. If I suddenly got 2 million unsorted images--I think I'd start pulling my hair out before I started trying to sort them. ;-)

But also, most of my images and video clips come from USENET newsgroups, where the subject matter is already fairly granular--for example, if I download a bunch of files from alt.binaries.pictures.erotica.facials , almost all of those files will belong in the Facials directory, and the ones that don't are either spam to be deleted or belong in the Blowjobs directory. Getting files from USENET groups instead of websites therefore provides a level of "pre-sorting" based on the subjects of the newsgroups I'm downloading from.

This of course provides a great number of duplicates over time, particularly since the same set of images could be suitable for several different directories; the same set of images of a black girl getting gangbanged and swallowing cum could potentially belong in either of these directories in my filesystem: BlackBitches, Cum\Facials, GroupSex\Gangbangs. In general I categorize according to a quick assssment of which aspect of the series is dominant--is the most attractive theme the hot black woman, the cum, or the group sex? If a series is exceptionally worthy, I might put links to those files in other directories related to the subject matter. Of course, much of the time, I just dump the files wherever they seem appropriate at a quick glance of representative files.

To eliminate duplicates and free up disk space, there are several applications that can be run to make quick CRCs of each file and compare them all, then make a list of duplicates to delete. The Win32 image viewer ACDSee comes with such functionality. I hadn't run it for a long time when I ran it last week, and so I freed up over 2 GB of disk space by eliminating duplicates. There are other such apps like TheChecker, DeDuper, etc., and I'm sure they have their Linux counterparts.

In any event, I have an actual life, so I do get way behind. Right now my Unsorted directories contain 104,792 files. :-( When I get the free time and have nothing better to do, I have a marathon porn-sorting session to cut into the backlog.

Database-driven filesystems are on the horizon, so that should provide an even greater opportunity to easily categoize files according to multiple traits.