- when launching, leave plenty of room in the explanation for huge later improvements:
- Apple can have any kind of magic behind the scenes, but if this is open it will be less possible to do so.
- maybe leave the base source open, but the web/app front end a bit different (it would have caching, knowledge acquisition, etc)
- its ok. Don't try to be Apple. Visible Magic (sidemind) vs Hidden Magic (apple).
- ! Who is the better magician, the one who must obscure his tricks, or the one who can leave everything in plain sight?
- what can we do that apple/siri won't/can't
- better interaction between users
- simple custom API
- how to leverage network effect
- share my music library with Luke
- getting people contributing to the language is I think the most important goal right now.
- make it easy to contribute
- allow people to contribute small bits to existing apps/domains
- if someone writes an app, and then someone else modifies the domain, who controls if it goes in or not? wiki style?
- allow people to add apps
- allow people to vote on apps which should be promoted to The General Intelligence
- create a proof of concept
- open source the compiler?
- the compiler could till use a bit of work and it might reveal the perceived 'magic'
- if we let other people create their own code (server side) have to deal with abuse. Could have some of the interpreting happen client side ...
- ! how to ensure quality?
- what are the standards to use?
- when to release?
- show an example of a decently polished app
- if apps are paid apps, how do you interface that with the user
- that command is available for $1, would you like to buy it?
- license to google/android
- perhaps consider 'hiring' people who I can work closely with and ensure quality and a clean interface
- android app
- say $5 base price
- (one time purchase, doesn't bring in much money for 'app store' like distribution pay)
- what if we say that it costs $5/hr (which gets about 3600 queries (2 years @ 5 queries per day))
- what about: $3/mo ($30/yr) in order to get 'cutting edge' functionality
- then with 10000 users => $30k/mo => half for costs, half to developers - algorithmicly determine who gets paid based on what people are asking - you can't set your own price
- later on, if there is a desire for special case languages, like for a particular business, then that can be dealt with in a different way
- for now, there is just one system
- will need to provide some kind of feedback system
- could also use kickstarter campaign to get the first stash of cash
- kickstarter money could buy:
- credits
- discounted voting points
- bounty points
- kickstarter kinda requires that people trust the idea will work ...
- kickstarter money could buy:
- with this system, there is still the possibility of fragmented tools. For example, "set the timer for xyz" and "set a timer for xyz" might wind up having different apps, which would confuse the user experience.
- maybe if two implementations overlap, this could be detected and one would be chosen, through beta testers or through a centralized system.
- centralized systems might not be easy to implement early on, but they could be strived for
- say $5 base price
- 'always on' on a computer (with a room microphone)
- website demo
- trueknowledge already does similar things ... though I think, not as intelligently :)
easy money:
- Android/Chrome app - http://www.sitepoint.com/html5-speech-input-fields/?
- pay for a webapp?
- pay for apis - won't last too long ...
- pay > $10 and get lifetime support
Not a business, but a neat way to get a very basic MVP: provide access on the web or in libraries for simple 'universal' parsers such as for parsing date/times, etc.
Also, something that would be interesting would be a pay model where users pay extra to have humans lookup and code the solutions that the algorithm can't yet handle
Is siri type approach the best way - perhaps some intigration with google ... do you have to become huge and powerful to do large things ? wikipedia doesn't, craigslist doesn't, we aren't quite to open free distributed manufacture of phones ... Maybe it isn't the right battle to be fighting? How can the right battle become core to the company though? Can be less direct ... I can try doing it with money made from selling the tech ... or w/e
Other related technology:
- Cogito: though they look pretty tame. Using expert systems and 'semantic' technologies, but are only bragging about advertisement placements, article key sentence finding and topic analysis. Nothing too too semantic.
How to limit growth or otherwise make sure the impression is given that axpress works in many situations?
I want axpress to learn, but it will take users asking the computer to do something it doesn't know how to do in order for that to happen. A few ways we could go about doing this, though most of them are pretty far down the line. How can it be released in a way that gives developers time to add functionality so that the ecosystem is good when the users get to it?
Is there a monetization strategy that I am excited about?
Possible monetization strategies:
- As a web service
- ads (surrounding content)
- ads (in content)
- prefered namespace order (pay to have images with tags bound to the flickr api)
- I'd rather have this be decided via some collaborative model
- pay to write the code for your API
- pay to add functionality to the code which you would like to see done
- ask amazon to pay if traffic is being redirected to them. don't redirect to amazon unless that is what is being looked for though.
I want to avoid allowing money to influence the service/software/experience in any way
+ are there exceptions to this rule?
What is a fair way to be compensated for this work?
Would people buy this as a solution to automated anything? A business could use it.
Can this logic and memory system garden?
What are the steps required?
What is the going rate for doing charity work? I'll pay you in that much free food to do it. because I have extra.
by charity I mean doing any service for free. for other people. I can chose who I want to give away my food to, and will naturally start out with friends. In the beginning, ideally to other people who would like to grow food to give away for free. I like the idea of gradually growing the number of people who are growing food to give away as possible. Ideally, to the point that everyone has food.
It would also be nice to make enough money to own a building and let other people pay me back by buying their own land? aka will lease
we don't need a socialist government to tell us how to be socialist!
is the goal to not have to work? Be able to give away programming, food, music, whatever?
You still have to buy the land, and the utilities, Offer an exchange: money for goods. spend all money on aquiring more land (so as to avoid requiring money, if it can't be gotten for free, sell extra and pay for it)
assumption:
own land (have no expenses) (or rather, have a surplus of money) growing food ? is this just a collective $ pot? pay for all of the people who are giving provide services exchangeable for USD
trade USD for services not available for free
start co-ops where there is no profit making. Or the hours spent at one co-op is redeemable (with an exchange rate) at other co-ops.
hours spent at the co-op are redeemable for service at the co-op. This means you can trade it for USD with people who want it. Could list it on a computer similar to an atm. Could set a default sell price (whatever the median is) or set your own. If someone wants to come into the restaurant, they can pay someone else to work for them. Whatever the going rate is for how much people are willing to work an hour there for. What about the managers? (they take the least out of the business as they can so they can pay people more for the work they are doing, and get better business (undercut)) The fewer managers and workers we can make, the better off we'll be.
There are initial expenses like the purchase of land, and there are service expenses, like gardening.
If we own the land we live on. And we grow the food we need, We can give away extra food (money to allow for others to join (aka buy/own more homes/land)
In the middle of nowhere, if you build on the land you can live on it. What more can we give away? (without having to sell?) If we can give away information, we can never be destroyed (aka the ability to share ideas with one another, the ability to copy and paste to each other, making free free)
give away an operating system. (that is the best)
I'd like to help continue the open source movement. I can do this most effectively if I can pay for myself to have this 'free' time.
semantic search.
With a natural language parser similar to english, can a semantic search work?
TODO: make work for one domain to show that it is possible. The next domain will be easier.
requires an implicit yes that was what I wanted or a no that wasn't what I wanted.
make a better search, and do it for free.
- requires p2p networking smarts ... or at least figure out a way to share the information with friends
we need free government internet. (or free private internet) public internet backbone could be payed for by the government and the rest through private funding (donations) (the library has internet, what if a city could just pay to be an ISP?) That isn't too much here pretty soon.