Kuro5hin.org: technology and culture, from the trenches
create account | help/FAQ | contact | links | search | IRC | site news
[ Everything | Diaries | Technology | Science | Culture | Politics | Media | News | Internet | Op-Ed | Fiction | Meta | MLP ]
We need your support: buy an ad | premium membership

[P]
The future of personal information managers

By maleficent in Technology
Thu Apr 05, 2001 at 08:28:05 AM EST
Tags: Technology (all tags)
Technology

Lately, I've become frustrated with the current choices in publicly-available personal information managers and have resorted to creating my own home-brew alternatives. The reason for this is that the current breeds of personal information managers simply do not fulfill the needs I have. How can we make them better, not just for me, but for all of us?


Over the last six to twelve months, I've used a number of the currently available PIMs in order to manage the great deal of information I have to deal with on a daily basis: seemingly bottomless to-do lists, random scraps of information, an enormous address book, schedules, and so forth. I tried a variety of products; among them were Evolution; Info Select; PIM Pro; Microsoft Outlook, Outlook Mobile Manager, and Schedule+ (yes, as a PIM, mostly for scheduling and Office tie-ins); Pocket PC running Windows CE 2.0; and a Handspring Visor Platinum running Palm OS 3.5.2. This essentially covered the gamut of non-trivial PIMs that I could come across using searches and word-of-mouth recommendations. I'm sure that there are many more that I have not tried that are non-trivial and recommendations are always welcome, especially if they begin to fulfill the goals I mention below. I've wound up writing programs for myself to manage my information; they don't work as well as what I describe below, but they do function for me.

All of the programs discussed above (besides my home brews) did a wonderful job of managing information; where they failed is managing my information. The distinction is a fine one, but a vital one; even books on the topic fail to make the distinction.

So, let's look at this distinction. Modern PIMs seem to take two approaches to solving the problem of managing information. The first, most popular, approach is the idea of making your information conform to their structures. You enter data into the fields already created by the software and search it using the methods they provide. The alternate approach, which Info Select used, was that all information is equal. It essentially treated all my information as bits and pieces, and searches would search everything and return all relevant results.

What's missing? A place in the middle in which software eventually conforms directly to the needs and uses of the end user.

Let's take a few examples to see what I am talking about. The first one I'll use is that of the task list. The usual task list takes in brief descriptions of the activity, puts it into a myriad of user-defined categories, and allows one to sort based on due date, category, importance, and so forth. A few such utilities allow sorting on additional areas of relevance, such as progress and so forth.

This method is very frustrating for me, as I'm often required to do several sorts in order to figure out what I should do next. The categories have different degrees of relevance to me and so I often have to sort by all the categories in order to figure out what is most useful at the moment.

A fantastic solution to this that I have yet to see implemented is a more intelligent task list. Such a task list would gradually adapt to the working style of the individual. At first it would function like a normal task list, as it has no information about your behavior. Over a period of time, though, it would see what importance you place on a variety of categories and would then adjust itself to list things based on your preferences. You could still sort by the traditional categories, but the default sort would take into account what fields matter most to you.

Many PIMs today can do this to a certain degree, as you can choose several layers of sorting. The difference is that in an intelligent task list, it would learn about your regular methods of completing and working on tasks, so it would know how important each field is to you, even to the point of determining at what value level a particular field takes precedence over another. For instance, I might be neurotic about getting everything done that needs to be done within the next two days at the expense of everything else, unless there was a high priority matter involving my significant other in the next week. Given time, the list would automatically learn to place the matter involving my significant other at the top of the list ahead of other matters, given repeated behavior.

A similar example could be given for each type of personal information. An address book with effective soft linking that helps with searching would be useful, as would a schedule that automatically figures out what things you do regularly and fits them in automatically without you telling it to.

Of additional use would be linking between information types. If the program figured out that I get a haircut on a roughly monthly basis, it might schedule in a haircut on a roughly monthly basis a week or so beforehand, stick a notice to call my barber with the appropriate settings on my to-do list, and link both to the phone number of the barber. All of this happens (after a learning period) without me lifting a finger.

The usefulness of this theoretical piece of software seems to be boundless. The only downsides I can perceive are the possibility of lack of privacy, which is relevant in any piece of software, and the objections some might have for letting software make their decisions. This group, however, would be the group that would be pleased with much of the current software available.

Implementing this would be a great challenge, to be sure, as it would push the boundaries of AI programming. True "learning" software is just beginning to be used and, interestingly, some of the best learning software available at a reasonable price is game software such as Black and White, which is quite successfully built around the concept of components that learn as you progress through the game. Applications to personal information will most likely require some new approaches and techniques.

Is this a reasonable goal for personal information management software? I think that it is. While there are a great deal of technical issues to be worked out, the advances in processing power today and the gradual evolution and improvement of software over time will eventually lead us down a path that manages information more effectively than today.

Sponsors

Voxel dot net
o Managed Hosting
o VoxCAST Content Delivery
o Raw Infrastructure

Login

Poll
What PIM do you use?
o A Palm-based handheld 30%
o A WinCE-based handheld 2%
o Some version of Outlook 4%
o PIM Pro 0%
o Evolution 1%
o ink on the back of my hand 13%
o something else 16%
o nothing at all 30%

Votes: 84
Results | Other Polls

Related Links
o Evolution
o Info Select
o PIM Pro
o Microsoft Outlook
o Outlook Mobile Manager
o Schedule+
o Pocket PC
o Windows CE 2.0
o Handspring Visor Platinum
o Palm OS 3.5.2
o books on the topic
o Black and White
o components that learn
o Also by maleficent


Display: Sort:
The future of personal information managers | 19 comments (18 topical, 1 editorial, 0 hidden)
Palm OS Todo list (3.75 / 4) (#2)
by scorbett on Wed Apr 04, 2001 at 05:31:15 PM EST

I used to use the Palm OS Todo list on my Palm III (before it died an untimely death) and I found it to be quite satisfactory. You could define arbitrary categories to place items in, assign each item a priority, and then display them on a category by category basis or display all at once, ordered by priority. I liked the fact that completed items were hidden, not deleted, so you could review how much you got done that day/week/month/whatever.

The only thing I didn't like was the fact that you couldn't really fit much into the short description, but this is more of a limitation of the Palm display and not the software. All in all I'd probably still be using it if I still had my Palm.

GIGO (4.00 / 3) (#3)
by Paradocis on Wed Apr 04, 2001 at 05:55:03 PM EST

Almost all of the features you mention in this article are already present in alot of the software you mentioned. You just have to tinker with the settings and configuration. I don't think it's reasonable to expect any peice of software to perfectly organize your life without being properly configured. Your problem sounds more like a Garbage In Garbage Out situation.


-=<Paradocis>=-
+++++++++++++++++++++
"El sueo de la razon produce monstruos." -Goya
+++++++++++++++++++++


Configuration... (4.00 / 1) (#4)
by maleficent on Wed Apr 04, 2001 at 05:58:36 PM EST

That's just the point. You use a PIM to save time. Having to perform a great deal of configuration works against that. The program should auto-configure to your needs.

It's also kind of insulting to see you refer to my personal information as garbage. I don't believe that was your point (you were referring to info in general, I believe), but it came out that way.



[ Parent ]
It's more like: (none / 0) (#17)
by ramses0 on Thu Apr 05, 2001 at 09:09:44 PM EST

...if you don't give the PIM accurate information (ie: Garbage In) you can't expect the PIM to give back accurate information (Garbage Out).

Just a metaphor, nothing personal. Being the astute reader that I am, I know that you want the PIM to capture your normal product usage (which is currently ignored by many PIM's), and slowly mold itself to your usage habits.

This is a good thing. It's just really difficult for programmers to do without some sort of framework.

--Robert
[ rate all comments , for great justice | sell.com ]
[ Parent ]

I didn't mean your personal info. (none / 0) (#18)
by Paradocis on Fri Apr 06, 2001 at 12:13:41 PM EST

I meant that if you put in garbage for configuration data, your output will be garbage, not your personal info. As for the PIM saving you time by configuring itself: it takes maybe an hour when you first run a normal PIM program (if you know what you're doing) to get a very complex and sophisticated configuration up and running that should be more than flexible enough, and then you don't have to worry about some automatically updating configuration screwing it all up. If you've set things up well, it's done, and it should be very rare that you would need to change anything. For that matter, many PIMs are highly extensible and will allow you to write plugins to add even more functionality if you need it.

I just don't see why an automatically configuring PIM would be desirable or even terribly useful. If you configure anything yourself, it's liable to overwrite it, and the (extra) code bloat would be spectacular.


-=<Paradocis>=-
+++++++++++++++++++++
"El sueo de la razon produce monstruos." -Goya
+++++++++++++++++++++


[ Parent ]
Market/Requirements a poor fit (4.00 / 3) (#5)
by babylago on Wed Apr 04, 2001 at 06:00:13 PM EST

I've been teaching myself XML today, so bear with me. If I understand correctly, your issue is not so much with methods of data storage, but methods of data retrieval and display. You want the retrieval, display, and possibly new bits to be stored, based on a continuing analysis of the data already stored.

Technically, this shouldn't be all that difficult. On the MS side, build a SQL database that holds the data, stored procedures for each of your analytical rules, and a dynamic front end that looks at the current priorities and constructs the display accordingly in an HTML or similar format. In some sense, this is what CRM systems do: run ongoing analytics based on rules and make recommendations and suggestions for a given context and time frame.

The trick is turning this powerful implementation into a consumer-level product. With my knowledge of current technology, I would rate this as unlikely, unless someone wanted to do an Access-type backend with VB rules instead of MSSQL/Transact-SQL stored procedures. You could certainly do this in *nix with the available tools, but it's not a good market fit with the people who want their information and interface prepared for them.

So, as I see it, you're down to a cost/benefit analysis that really doesn't work out with current technology and market demand. Sorry.

---
[ Blog | Hunnh ]

Changing rules as well... (none / 0) (#6)
by maleficent on Wed Apr 04, 2001 at 06:03:57 PM EST

Obviously, the nature of the rules governing the information relations would change over time as well, but I agree that information storage and retrieval will have to take a step forward, too. It will, of course, require an improvement in a lot of current software to pull it off; such software with ubiquitous usefulness should be a goal, I think.



[ Parent ]
but... (4.00 / 1) (#7)
by _Quinn on Wed Apr 04, 2001 at 08:52:24 PM EST

how will you define the data structure for the

`SQL database that holds the data, stored procedures for each of your analytical rules'

? I would think storage in XML and interface by XSL(T)/XUL would be a better way to go, because the structure of the data will be not uniform across time, and the semantics are more flexible.

   The problem with the article's proposal is feedback to the learning system. I can't (after a little bit of thought) come up with a system which would offer enough feedback for both fast and good learning without being as annoying and obtrusive (initially) as doing an extended (extensive) configuration (customization) in a normal (or super-powered, whatever) PIM.

-_Quinn
Reality Maintenance Group, Silver City Construction Co., Ltd.
[ Parent ]
What I use (3.50 / 2) (#8)
by cameldrv on Thu Apr 05, 2001 at 12:07:33 AM EST

I use the Franklin Planner software for windows, and a palm pilot running the Franklin apps. The primary thing I use it for is the to do list. The way Franklin works is that you have a "master task list" which holds things which are long-term. You also have a daily todo list which holds things which are supposed to be done on that day, depending on the priority. The idea is that you sit down at the computer and decide what things you would like to do each day or each week and schedule them. Personally I'm a pretty disorganized person and I certainly don't complete all of the things on my daily list each day. However, just the act of going over the master list and finding things I really need to do, and then scheduling them helps me wrap my mind around things. You can forward tasks from day to day, so if you don't get it done you can move it back to the master list, or forward it a day or two. I find this very useful, as it forces me to acknowledge what I don't accomplish in a given day.

With regards to more automated/AI approaches, one product I found was Life Balance, which allows you to set priorities in different areas of your life, monitor how you spend your time, and suggest activities for you to do at the present moment. It even will deal with different physical locations and suggest activities you can perform when there. For example, it will suggest different things when you are at work and when you are at home. I found the system to be very cool. The major problem I had with it was that there's no desktop software, and in order to efficiently enter a lot of tasks and look at analysis, the palm really isn't efficient. Supposedly they are working on this though.

my personal information manager (3.28 / 7) (#9)
by eLuddite on Thu Apr 05, 2001 at 01:02:24 AM EST

cons:
  • there's a bug in the way it sorts tasks according to importance
  • the map algorithm is totally b0rken
  • promiscuous network interface - very chatty on the wire
  • random core dumps
  • license locks you in with exclusionary and non compete clauses.

    pros:

  • 3d interface (and how!), more skins than themes.org
  • natural language interface
  • more easter eggs than the Vatican in April

    Although it comes in many different models, it's generic name is wife.

    ---
    God hates human rights.

  • outlining programs (4.00 / 1) (#10)
    by sjwillis on Thu Apr 05, 2001 at 09:34:01 AM EST

    as much as it's linear approach is somewhat flawed, i'm a big fan of outlining-type todo lists. my current choice for palm outliner is arranger--http://www.olivebr.com/pilot.htm. being able to manipulate my todo list in like fashion on my linux desktop and sync it back to my palm would solve pretty much all my pim problems.

    Check out The Brain (3.00 / 1) (#11)
    by argent on Thu Apr 05, 2001 at 11:29:55 AM EST

    One piece of software that I've played around with in the past was the brain. (www.thebrain.com)It allowed you to link programs, URL's, and documents according to how you thought of them and create cross links, etc. I'm prety sure that it's Windows only, but it's pretty cool.


    argent


    cd /pub more Beer
    Laugh at me if you will.. (3.83 / 6) (#12)
    by Zukov on Thu Apr 05, 2001 at 12:05:19 PM EST

    But I use a combination of 3 things.

    First, my memory.. I tells me what to do next and gives me a list of frequently used phone numbers, as well as regularly recurring meetings. Deliberate training of my memory greately increases the capacity.

    Second, a small book made of a substance called "paper". This book has mostly blank pages, with dates and times pre-printed on them. When an isolated event is to happen far in the future, I write (smear black "ink" on the "paper" in predefined patterns) in the appropriate place. An area of this book contains a slip of paper which has a superset of needed phone numbers and information. A great advantage of this book is that the record is really permanent, and every year I get a new book, (cost - 72 cents) putting the old one in a file cabinet. I have had need to look up a date from several years back, and the book allows me to flip thru and find what I want. I use waterproof ink, so immersion in water does not destroy the data. The book can be sat on, driven over by a car, give electric shocks, etc, and the data stays intact.

    The third item is a huge text file in the system into which I put random data, phone numbers, contact info, and key words. The entry of data is really free form, and can include headers from mail, etc.

    When I want to find something I grep this file by keyword, portion of name, area code prefix, whatever. I keep backups of this text file on -gasp- a floppy (well, I keep more than one backup, but the file is less than a floppy full).

    I realize this all sounds hopelessly old-fashioned, and I admit that PIMS are cool/neat/can_browse_the_web/etc. I question how much better off people are when you have to spend time fussing with getting their data into/outof/synchronized or lost due to system failure with a PIM.

    ȶ H (^

    Yes, I have just bumbled upon Gnome Character Map. Please ! me.

    Target Audience (4.00 / 1) (#13)
    by ucblockhead on Thu Apr 05, 2001 at 12:21:56 PM EST

    The biggest problem I've had is that my needs aren't quite the same as the sort of person PIMs are typically targetted towards. Sounds to me like you've got the same problem.

    Most PIMs seemed to be geared either towards sales types, or managerial types. I've no problem with that, but someone who programs all day has very different needs. Like you, it is the task list that always seems the biggest problem. They typically just aren't powerful enough. Yet the calendars all always overkill for what I do. I typically end up with maybe three entries each week, one of which says "fill out timesheet".

    Given the types most likely to buy PIMs, I don't see this changing in the near future.
    -----------------------
    This is k5. We're all tools - duxup

    Learning and agents (5.00 / 1) (#14)
    by kellan on Thu Apr 05, 2001 at 05:52:08 PM EST

    To me it doesn't sound like your looking for a PIM so much as an intelligent agent who manages your PIM. Learning, pattern recognition, intelligent scheduling, that all falls solidly into the agent field fo AI research.

    Look for anything thats come out of Pattie Maes' group at the Media Lab in the last 5 years, including Webhound, HOMR/Firefly, Open Ratings, Mobile Agents for Routing Discovery, and Maxims.

    http://agents.www.media.mit.edu/groups/agents/
    http://agents.www.media.mit.edu/groups/agents/projects/

    A quick look reveals something called:

    Calendar Agent (1992-1993)
    by Robyn Kozierok, Pattie Maes
    The Calendar Agent learns a user's calendar scheduling rules and preferences by observing the user's actions and receiving direct feedback.

    A classic in the field is Yenta:
    http://yenta.www.media.mit.edu/projects/Yenta/

    An agent with an emacs interface is Remberance Agents
    http://rhodes.www.media.mit.edu/people/rhodes/RA/

    Also a project that was recently GPL'ed and looks promising (not that I've looked at it all that closely) is:
    http://www.logilab.org/narval/

    I'm personally looking for a different feature that is missing from my PIMs, but I'll post a different message about that.

    Kellan

    I feel your pain (3.00 / 1) (#15)
    by Big Dave Diode on Thu Apr 05, 2001 at 05:54:43 PM EST

    I looked long and hard for a system to keep track of my personal info that wouldn't lock me into some proprietary format and that worked exactly the way I wanted it to. I finally figured out that the best way to handle my data was to do it myself, so I wrote a web interface on top of a MySQL db and got myself a web hosting account. I'll post the app to sourceforge as soon as I clean up the code to the point where I'm not embarassed for other people to see it :-).

    I know this is kind of a smartass answer, but this solution worked well for me!

    PIM as collection of services (none / 0) (#16)
    by kellan on Thu Apr 05, 2001 at 06:33:36 PM EST

    I also find the current crop of PIMs uncompelling, but for a very different reason. I'm not interested in something which can figure out I need a haircut and schedule an appointment for me, a problem whose tractability I have some serious doubts about.

    My problem is that I find PIMs are too far removed from my life, the calendar/todo list/address book is way over "there", and I'm here, and to keep it relevant to my life, I have to pour almost slavish devotion into it. (Maybe meeting a friend for coffee Thursday, better pencil in a tenative meeting) I think this is because the conduits that we have to get data into your PIMs are too rigid and unaccessible. PDAs have made this a little better, I can now run into someone on the street, and scribble in that phone number, or coffee date, without having to hold it in memory until I'm sitting at my computer again, with my PIM running. Still, that is only one of many sources of data, and unless I'm religious about it most of them slip by me.

    The solution is to have your PIM be able to listen on your channels of communication. Perhaps this sounds like maleficent proposal, but I'm not asking mine to be intelligent or to operate without the metadata. If I'm exchanging email w/ somebody there should be a trivial manner by which I can dump that email into the PIM can add them to the address book, linked with any events, or tasks that were being discussed in the email. Similarily messaging. And all the rhetoric around web services right now could really lead to some neat convergences. The classic example if the airline reservations get sucked directly onto your calendar, and then if your flight gets cancelled you get a note about it. However all of us on k5 know there are a variety of other ways you interact with the web then booking airline tickets, and I think a PIM will need to be able to address those.

    I also think PIMs fail by trying to be all encompassing. I'm not going to use Outlook (except when I have to for corporate reasons) to send and recieve email, just so that I can have a calendar that talks both to my email, and to the people in my scheduling community. I want to be able to move back and forth between pine for email, a website for a calendar, jabber for IM, and then sometimes, do all that through my palm pilot.

    And when something get updated, or I need to be reminded about a task, I want the PIM to come find me, via email, or IM, or whatever.

    kellan

    you need a personal secretary (4.00 / 1) (#19)
    by kubalaa on Fri Apr 06, 2001 at 05:49:12 PM EST

    Argh, these kind of ideas make me really mad, because I don't think people have any concept of what you're really suggesting. Let's look at the assumptions:
    1. Your PIM has access to enough information to make these extrapolations.

      That means that, in order to know that you go to the barber once a month, it has to be told somehow. So either you type in every month (until it "learns") that you want to go to the barber, or you have it hooked up to some complicated AI and camera system which monitors your every move. Either way, you're giving up a lot of privacy, or you're spending a lot of time entering useless data for it to analyze.

    2. Your schedule can be defined with some regular mathematical precision.

      Well, let's look at it this way; who decides when you do things? You do. The most accurate model possible of your behaviour is already running inside your head. Any other model is inherently less accurate. Obviously you don't expect your PDA to be able to predict that next Thursday afternoon you'll spontaneously go for a walk in the park. Or that your client will call and ask you to lunch tomorrow. But it's just these kinds of things that we need PIMs for. You don't need it to tell you to get a haircut, or brush your teeth, or go to church every Sunday; you need it to keep track of things which happen so rarely or irregularly that the incredible predictive engine that is your brain can't keep track of them. The question is, do you really do so many activities with mathematical regularity that you can't keep track of them yourself? If you're that predictable, save yourself some money, sit down with a calendar, and work out for yourself your schedule for the next 5 years.

    3. With this information, your PIM is capable of making these extrapolations.

      Humans tend to calibrate themselves to certain regular measures, both artificial---hours, days, weeks, months, years---and natural: position of sun, length of day, temperature, moon phase, tilt of earth, age, location, weather, etc. Assuming that the behaviour you want your PIM to predict can be described as some function of these quantities, how well can it be done? Going to the barber once a month is pretty easy; it'd take your PIM probably about 5 points before it could extrapolate with reasonable certainty that it's a regular function. (You really think entering 5 visits to the barber shop and hoping your PDA "gets it" is better than saying "Repeat: every month"?)

      But I imagine these simple cases aren't what you're interested in, so let's look at your idea of intelligent prioritization. Assume 5 variables: importance, date due, completion, category, and type of task. (This makes the strange assumption that most tasks can even be well-described by these 5 variables; there's a reason most people label and describe their to-dos) How many points do you think it takes to extrapolate with high probability an irregular 5 variable function? A lot. How hard do you think it is? Very hard. How high is the probability that your PIM will find the "correct" function before you get tired of it and go hire a personal secretary instead? Very low.

    4. You want your PIM to make these kinds of decisions.

      Well, I'll tie this in to my previous point: if your priorities are simple to decipher, you'd save yourself work by telling them to the PIM directly. What you're implying is that you are less capable of understanding your own priorities than a glorified calculator. If you need your PIM to tell you when your S.O. is more important than your work, you've got problems. If you KNOW, to use your example, that your significant other always comes first, then why don't you just TELL the PIM so, by placing a high weight on the "significant other" category? Why does it have to guess this? If you don't understand what's going on in your own life, is your PIM really going to help you?

      The other implication is that you'd become 100% dependent upon your PIM. You're basically relegating all decision-making to it. Want to watch "the game" next week? Better tell your PIM, or else it might automagically schedule a dentist's appointment instead (since you've been entering them faithfully for the past 5 years, it's finally figured out that you go to the dentist about every 9 months). Or what about that important meeting with your new client? You'll have to hope your PIM somehow knows about your complete financial, social, and psychological situation enough to deduce why and how important that meeting is.

      One last point in this vein. We have to assume that the AI behind these calculations is too advanced to be easily understood (after all, if you could understand it's predictions you wouldn't need it to make them, you could just make them yourself). Using software with what amounts to emergent behaviour is like using closed-source; you don't really know what's going on, so you can never really trust it, you don't know its limitations, and you can't trouble-shoot when it breaks. Say you go to the gym Tue and Thu, but you get sick for a week and don't go. What if your PIM "intelligently" comes up with this complicated rule to explain your behaviour, like "gym every prime day of the month plus the third thursday except during the winter soltice on leap years or after seeing the dentist." Well, we can hope it's smarter than that, but the point is you'd never know what rules it's using until one day it mysteriously fails to schedule your visit to the gym and sends you to get your legs waxed instead.

    Here's the prime failing of your idea; it assumes the computer is smarter than the user. That's why many people hate Windows, because it makes the same assumption, and it's wrong. And when the computer makes a stupid mistake, which it inevitably does, good luck winding your way through all the advanced logic to fix it. Until the point when AI has truly surpassed humans, you are much better off with a system which does exactly what you tell it to, no more and no less.

    Now, it's possible that I've over-estimated the amount of work you expect your PIM to do. However, I'd say that, unless it's more intelligent than you, any "intelligence" it has is going to fit in one of three categories:

    • Passive: like auto-completion in IE., In my mind this is the best and ONLY role for AI in current computing. In PIM terms this is like automatically using the current day when you create an event, but making it easy to move it to a different day. But this is quite clearly not what you're describing, so it's irrelevant.
    • Active, but simple: like how Excel "intelligently" adjusts formulas for you if you move or copy them But, the rules it uses to do so are still simple enough that you always know what it will do, and it's easy to work around it. This can hardly be described as intelligence, since the program is still 100% predictable; it's plain command-response. In PIM terms, this is like how Palm DateBook creates a new event when you start writing. This is also clearly not what you want
    • Active, but complex: where the program does things for you that are complicated enough that it's no longer a simple command-response interaction. While this is the most useful kind of intelligence, since it requires nothing of the user, it's also the most dangerous. It will inevitably fail, and when it does, it's too complicated to fix or work around. The only place this has is in non-critical applications, or in the passive role described above. I might trust an AI agent to recommend products to me, I'd never trust it to go out and buy them.


    The future of personal information managers | 19 comments (18 topical, 1 editorial, 0 hidden)
    Display: Sort:

    kuro5hin.org

    [XML]
    All trademarks and copyrights on this page are owned by their respective companies. The Rest 2000 - Present Kuro5hin.org Inc.
    See our legalese page for copyright policies. Please also read our Privacy Policy.
    Kuro5hin.org is powered by Free Software, including Apache, Perl, and Linux, The Scoop Engine that runs this site is freely available, under the terms of the GPL.
    Need some help? Email help@kuro5hin.org.
    My heart's the long stairs.

    Powered by Scoop create account | help/FAQ | mission | links | search | IRC | YOU choose the stories!