Jeff Fitzmyers: MS in Hydrology. Knowledge base: systems, design, money, trading, psychology, botany, healthcare. Creator of the only real Life Cycle Of Trauma on the net.
Insight on my process.
2014 November 22
I have been catching up with current events in AI. Result: Intelligence / value / ROI is more the connections than in the nodes. AI does not currently need to create a better silicon brain. Just connect the existing 20+ billion carbon and silicon based brains correctly.
Also, I have been investigating how to apply Alex Wissner-Gross‘ work on intelligence to Castpoints. Currently, people would just pick from the top ranking ROIs tasks and do that. But what if an event has more people than money? A whole different set of tasks based on scarce money and abundant people would be a more intelligent choice. Also, since we know the history of how and when successful events allocate resources, we can easily suggest to events when to make new roles, add people, and allocate resources in general. And of course some of this can be conditionally automated for an event. Pretty cool!
Seriously!?!? If true, most poor decisions are history.
2014 November 10
I really wish I had “Castpointed” the making of Castpoints! At first I tried to make a demo, but it did not have enough interactivity to show what’s going on. And also I needed specificity in the backend to organize and precipitate the ideas. So then I tried to code it myself, which is great, but now I’m low on savings and am spending a lot of time trying to raise money. Plus small changes could mean a lot of recoding.
Now I understand, the answer is “both”. Code the barest minimum, and demo out the rest. My focus should have been on me mapping it all out, so that I could precipitate the design, then replace demo with code.
I was overly involved in the trees, not the forest.
At my 2000 dot bomb job, where I was originally hired to make the internal CRM tool, the main application was a website that tried to match commercial service providers and customers. Window washing, paving, etc. The data base had 100+ tables and when printed out in small font would cover a whole wall. Castpoints matches anything with anything yet only has about 30 tables so far. That’s the power of abstraction, and KISS. But I have to keep deleting stuff I “think” is “needed”, which takes time.
2014 November 3
I have been working on this off and on for over 13 years. I have failed > 3 times.
Version 1: I paid a person to code it. He tok the retina and I did nothing.
V 2 I taught myself to code it but the design was overly complex and I got bogged down. This is when CP was just dealing with content. Thought I needed to code plugins for CMS. I’m amazed at how I didn’t just sit down and route around this perceived limitation.
3 Tried again. Since the content part was so hard, I addressed the goods and work part. The design still did not mesh together. Also, I really feared the gov would really not like it
3.5 Tried to ignore it and give up. But I couldn’t. Everywhere I look, I see how Castpoints can help billions of people. I often felt guilty that while I was fiddling around procrastinating, real people were really suffering.
4 Many breakthroughs via inspiration. Now I have a self consistent awesome design, but can’t explain it to anyone. In fact, rarely will someone understand 10% of the explanation. And I try desperately to talk about how CP will help the exact things they are interested in. Yet, they don’t get it. I still feel stupid. If I spent half the time fiddling around on actual coding, it would be done by now.
4.5 More insights, I realize CP is evolution codified. I have always refrained for saying CP will solve pretty much all problems. It sounds quite egotistical, and I did not know the exact hows and whys. But now I do. Its simple, just correct the original mistake. But still, I guess we are all too used to constant strife and suffering to.
I got frustrated with VC’s, Angel Investors, etc. Most say they want buzzwords like paradigm disruption, unconventional, willing take risks, etc., but all I get are rejection letters. Granted, my marketing is probably average at best 🙂 But really, how can I expect them to get it, when I have been working on this for 10 years and only recently got the whole picture?
I hear Silicon Valley is funding anything right now, so the fact I keep getting rejected is actually positive from a contrarian view point.
I traded emails with principle scientist at HP. He was very helpful! Basically he said code trumps explanation any day. So I am mostly working on the demo right now. It was also nice he labeled what I was feeling. “ won’t be easy for a system that’s this integrated.” “…the goals are admirable…” Yay! He really gets it! Wait, perhaps he is crazy too?!?!? 🙂
I just had an epiphany in the shower. (Typical place.) The question is how does one value a post? For a while I had it too complicated. There was a rating of content, presentation, selection of theme, and how much money its backed by. (This is embarrassing!) The goal was to separate the value of content from secondary aspects. (I used to have a complex icon that would denote all this, insurance, and context.) Tangent! Here are some of the icons. I can’t find the complex one.
I finally reduced this down to the rating (% of the post people found valuable to the person: 0-100% in steps of 10), and the amount of money people backed this up by. This is as far as I could go until I fully developed the classification system.
I was just pondering the difference between ROI and rating. Now that I had the support of the classification system, I realized the rating is redundant. All that’s needed is the backing money amount. So now the user goes from having to make 2 decisions, to one. And under the hood, CP will note the % of their capital used — this denotes how much people really find it valuable!
But now I need to go back and change code. Which is one reason why things take so long.
The overall vision is crystal clear, including some emergent properties. But the details of getting there are still being birthed. Since everything is so integrated, the best strategy has been working on one area for a bit, then moving to another area. Then I can go back to the first area and refine it. If I would allow myself to skip around more, I would get more done.
How’s this for one of the first tries are a comprehensive search? Its a joke 🙂 Worse, the categories were not auto configuring. But I had t start somewhere. That was just one of the first iterations. I’m on version 5 now.
I just found an old “history” post. Please don’t read it. Things have come a long way!
Our initial vision was an advanced, consumer friendly email marketing system. What we ended up doing was organizing aspects of the internet and putting a small price on contacting anyone.
Its really ironic to help create a site that revolves around advertising
and marketing because I don’t particularly like either. I believe good products
need little more than word of mouth and awareness advertising. A good product
or service will sell itself. However, the 17th kind of mustard does need to
be expertly marketed because it is not really adding value to the market place.
Now marketers find it difficult to convince consumers that the mustard (dress,
new car, latest software update) is important enough to spend money on. Since
logic will fail to convince people, emotions are used. “This (product/service)
will make you More Sexy, Live Longer, Look Smarter or Feel Better, if you give
us a small bit of money. They trade the consumer an illusion for money.
The sad thing is, this is not really a fulfilling relationship
for marketers, consumers or vendors. Valuable time is spent chasing and promoting
illusions. It is inefficient and expensive, and I doubt it will last long in
the deflationary environment of the internet. What people really seem to think
has value to them is a personally efficient combination of technology, fun,
community and low cost. The idea of Infofitz is to provide a structure for everyone
to get maximum value at minimum cost. How do you do this?
I think economics is primarily consumer driven. If a vendor offers
something that has little value, they will fight an uphill battle to get consumers
to buy. If a vendor offers something that has value, and consumers become aware
of it, it will sell. If it has great value, people will break down doors and
overlook most hassles to get it. Now, if the product or service is really new
and ahead of its time then vendors and marketers have to educate consumers.
When value is offered, and consumers understand the value, wallets are opened.
If the consumer can clearly state their interests, vendors will
be happy to fill the order—it is the path of least resistance; marketers
will be happy to facilitate the process. Wondering if a new product will go
over well? Test it in a few minutes for the price of a great meal.
How do you get the consumer to figure out what they are after?
Some know exactly what they want, others just don’t. So we make it cost something
and give everyone feedback. People pay attention when money is involved.
We wanted to reduce spam while increasing privacy and safety. While researching
the market, we started to collate some of the excellent information into a
portal for later reference and as Resources for the site. Our research showed
that although customers would be likely to sign up, Vendors might not. So
we set out to entice Vendors.
We did not want to have any advertising on the site because
we wanted to appeal to the people bored with banner ads and the download time
they take. But what if a company copied the web pages and subsidized Vendors
with advertising money. Well, we figured we would just preempt that whole
concept by letting people decide what they wanted. If consumers don’t mind
banner Ads, through the feedback mechanisms, they will be able to charge more
for receiving messages.
3/1/00 – Naming
We really wanted the site name InfoDepot.com. It rings and is short and clear.
But it was taken and our meager budget precluded purchasing it. We took a few
weeks to think of InfoTreasure.com. We like it because treasure reminds us of
treasure hunt which is fun and has fun rewards, but you have to work a little
at it. As much as we have tried to keep the site simple there are a huge amount
of choices. An update: We can’t use InfoTreasure because a guy called to say
it was his and a his company forgot to pay for it. He could prove it, so although
we are looking for another name, we will probably stick with Infofitz.com Another
update: Apparently his company folded because he sent $70 but I could not just
send him the password because it is the same password for Infofitz. I emailed
him and left voice mail twice with no reply. His company name is InfoTreasure
so that creates trademark issues for the name. I will sell it after trying to
contact him a few more times.
3/10/00 – Partner programs
We finally broke down and added the ability to become a Infofitz Partner. We
did not want to do this because it adds to the complexity of the site and we
have never liked selling to friends. But while adding to the Email Marketing
Guide we referenced a graph that showed how much better affiliation was compared
to banner ads. Well, we can’t get a good reputation by saying one thing and
doing another so now partnership is available. Of course, now that we have it,
we like it.
3/14/00 – Protecting children
Kids are on the net, unsupervised all the time. You simply have to design with
this in mind. That is why there is no adult content anywhere. At first we were
not sure how to handle adult content. We thought of sending in proof, working
with a third party, or having a separate section. We ditched the whole thing
because of the added complexity and responsibility. If someone wants that service
they can create it or they can put adult material on the distributed search
3/17/00 – Where and why the money flows
How much should Vendors be charged? We have been struggling with this for
a while. We tentatively settled on 10% mainly because it was in the middle
of the range we are playing with. 5% probably won’t cover costs; 20% might
whet the appetite for competitors. But Mary Modahl’s book Now or Never gave
us the clues to back up our number. She says people helping generate a sale
should get about 5% (our target for Partners), and taking orders and delivering
is good for 15%. Fulfilling products and handling returns is good for another
25%. Add all those up and the max we could charge is 45%. The minimum is 5%.
So how much to charge?
With Vendors and consumers already paying about 5% to partners
(that is about how much we should be paying for advertising anyway) and if Infofitz
charges 10% a flat rate, Vendors will be paying 15%. That is pretty good considering
Infofitz delivers the product and crunches all the numbers and ratings into
a valuable format. Infofitz also can make money form signing people up for the
micropayment systems. If we start to do well there, we can lower the flat rate.
The system is scalable, and the fact we don’t have to guess correctly how much
money to allocate. It all happens on a percent basis. As the system grows it
is able, by definition to support the growth (if it works).
Some consumers will not sign up for e-gold (they still can charge
a fee). This money will allow us to lower the flat rate to Vendor purchases
3/27/00 – We like competition
We saw the site YesMail.com today. We really like how they have the categories
set up. But the flexibility in the Infofitz system will serve us better. You
can make a profile for me just as easily as a friends for a birthday present.
It did get us thinking that we need to move the consumer’s fee selection from
the account area to the interest profile selection area. If someone was interested
in a high ticket item, they would need to charge more to keep the scammers out.
Yet, if they were interested in a low ticket item, they need to have a low Fee.
3/31/00 – Micropayments
I am having issues finding great micropayment systems. e-gold is excellent
and has been around longer than any others. I am defiantly using them but
I would like participants to have other options. X.com got very bad PR from
security problems. PayPal does not seem to be merchant friendly. The others
have either folded or are so new they have no track record. We are looking
into Beenz but have trouble getting information out of them.
Micropayments are a great thing for merchants because overhead
is reduced and transactions clear quickly. This is because real money, not
credit, is in accounts. That is the challenge: the consumer has to sign up
ahead of time and fund an account. But once the account is funded there are
no credit checks, no risk management and no personal information needed. The
account either has money to spend or doesn’t. Merchant don’t have to worry
about charge backs, stolen credit cards or who your are—dramatically
reducing overhead and accounts payable.
Infofitz can’t take credit cards because the money goes almost
instantly to consumers. If the payer did a charge back for $5,000 we can’t take
that money back from 500 customers. So we have a $5,000 cash flow problem. No
4/16/00 – More on micropayments
Assuming the net is deflationary (higher apparent supply drives prices lower),
and that consumers and Vendors always want high confidence, low cost transaction
systems: Credit cards and ecommerce doesn’t mix. Credit means fraud which
raises transaction costs. Credit is hard to get if not a USA citizen – vastly
reducing the consumer base. One also has to deal with currency conversions.
Solutions: 1) Checking or savings account debit transactions.
How willing are people going to be giving up their info on the web? There
is very little ‘insurance’. Apparently the current debit system is inefficient
and costly. 2) Digital debit value that happens quickly and cheaply. What
is a valuable thing that almost everyone in the world agrees has some value:
gold. This is what e-gold is and I hope is doesn’t go the way of Beta Tapes
because we have not been able to do business with any other micropayment system.
They simply don’t communicate, cost too much or are not merchant friendly.
We think the major credit card companies just shot themselves
in the foot by raising transaction fees as a way to reduce fraud. They are begging
for competition and ignoring small Vendors. Why not issue another card for internet
debits: the “International Debit Card”? We don’t see a good reason
why not (although we are not experts)—they have a system in place already.
5/3/00 – Not live yet, not even programed
We thought the site would have been finished 3 months ago. But as we thought
about the data we had to play with, we kept coming up with ideas on how to use
it. Users provide very little data. Yet by mixing and matching it together we
can generate a lot of very useful information.
Another thing slowing deployment is the last site version was
so confusing the programmers could not figure it out—and they are paid
to. We had to completely redesign the navigation and organization. The problem
originated when the site was 20 pages the navigation only needed to do a few
things. When the number of pages got over 100 this style of navigation did not
scale up, which created a mess. That is 1 thing we have really learned: plan
for easy, if not automatic, scalibility
Categories: Our research showed that basic users just want
a few things to click on while advanced users, people that have money on the
line want lots of ways to really describe an interest, and proof of their investment
return. Initially we were trying to design for both groups at once which lead
to a junky result. With the tree structure classification system we have designed,
both groups start generally and stop when they get detailed enough for their
5/5/00 – Categories
After more thought about the categories, we decided to make them open ended
so that we can catalog everything. Not just computer nerd items and shoes, but
stars, the plant kingdom, chemicals, anything (if we want and can do the catalog
maintenance). With a branching structure, if you just want to have a general
interest, that is fine. But if you really want to be specific you can by making
more selections. The complexity is hidden. This allows niche and specialty vendors
to display their products. Anyone can look for a book at Amazon. But where do
you go to find those little candle snuffer things? What are they called, technically?
If I sell those candle snuffer things how do I get the word out? Which search
engines do I list with?
5/20/00 – A list of problems Infofitz solves
- Out of date information
- Small email marketing budget
- Consumers not being able to search for a mix of cost, customer service,
location and product quality
- Consumer privacy while ‘shopping’
- Vendors can’t control what information about them and their products are
listed in search engines
- Potentially not being aware of new products in your field
- Can take up to 5 months to get listed in a search engine
- Many web pages not in search engines
- Poor search engine results
- Lack of confidence and little way to gain confidence while considering a
site or product
- Can’t search by any type of specification (how big, how heavy, how much,
- Small vendors lack resources to get good positioning in major search engines
- The wrong vendor information being shown in search results
6/28/00 – More usability challenges
The programmer said things, especially how you define a profile with the menus,
are way too complicated and the search method we were interested in was not
possible with the data base MySQL. So we had to make the site better. A different
data base way easy. We switched to PostgreSQL because we could do the searches
we want and it is more robust (but probably less speedy). This did not used
to be important when the site was first conceived – the data base was simple.
But now companies might have spent many hours creating request profiles and
we better not lose any information.
Going back to the search issue. Unlike standard search engines
that have to deal with misspellings, upper/lower case letters and little or
no context, we have everything clearly defined and so should be able to match
interests exactly. We took out the “motivation of interest function”
menu because people might get confused. Instead we are putting things like “clean”
or “transport” in context with the actual items. This will make more
branches in the interest tree, but then the user will not have to have a lot
of knowledge with where everything is.
7/4/00 – Defining our mission even more
Infofitz is all about being direct and using economics as a filter or traffic
cop. Everyone is in charge of their own information and motivated to do a good
job by economics. Now vendors and consumers can exactly describe, “Who,
what, where and how”. Feedback makes sure (I hope) everything is honest.
8/2/00 – Redefining our mission part II
“Infofitz is all about being direct” yada, yada. We are probably merging
with people who have been working on instant, secure and encrypted payment systems.
They will bring the site full circle. See something you want, just click on
buy and its on its way to you. Want to automatically get 10 things from the
highest ranking vender at the lowest cost each month or week, just say so. We
are all so excited about this we can’t sleep. Basic inventory, shipping and
volume discounts are now incorporated. We were a little worried at first with
a merger due to our business inexperience (and a certain persons gullibility
“Did you know the word gullible is not in the dictionary?” “No,
your kidding!?”). We were made aware of our loss of control. But we were
losing control anyway due to the work load and small staff. The group we are
hooking up with have a great reputation and so we ditched our friendly worries
for some potentially good synergy and exciting times. Tune in for future installment
to see if we blew it!
8/3/00 – If the site fails, you will know the real secret why:
To many flippin animals that need me to be the center of their attention. I
have a cat in my lap and a dog at my feet. If I get up – they get up. If I sit
down they sit. If I type, the cat wants to help. If I put them out, I get the
poor sad eyes.
8/19/00 – A move to open source?
We will run out of money soon. We did not expect to make things so complicated.
Mark Miller (Erights.org) was EXTREMELY
helpful, even after I said I couldn’t hire him. He said we were including too
much innovation in the first version. If we can’t go VC (I have not really tried)
then we will have to go open source. I meet a few people at the Linux expo and
asked about the technical issues. I need to follow up, but it seems it is best
to use Sourceforge the free, industry
standard (I think) for open source projects. I am reading up on it now. I can
be a vision guy but not (yet) the hard core tech guy because my programing education
is still too young and so can’t be respected by the open source community.
We did think of a way to reduce 80% of the interest profile complexity
while maintaining 80% of the precision. Good enough is good enough. All we need
is a clever, practical, flexible implementation plan. 🙂
The free software movement , I am learning is very cool. It is
a a gift culture. No that we are aware of it, it seems necessary to go this
way because Infofitz is based on free market principles and we would almost
be hypocrite to do otherwise unless otherwise had large benefits. I do worry
about the code forking, but less and less as I read more.
|Culture type||Based on||What determines social status||What they do||Examples|
|Gift culture||Abundance||What you give away||“You may not work to get reputation, but the reputation is a real
payment with consequences if you do the job well”; have fun
|Open source software, interactions among the very wealthy, show business|
|Exchange economy||Scarcity||Having control of things to use or trade||Allocate scarce goods through decentralized trade and voluntary cooperation;
|Command hierarchy||Scarcity||Coercive power||Usually a parasite on a larger economy of a deferent type; the larger
they get, the more brutal and inefficient they become; scales poorly
|Source: Eric S. Raymond, Homesteading
9/16/00 – Regrouping
We have a meeting with a group that might provide us with resources. I am taking
a class to learn SQL so I can do the data base myself. I would never have thought
it would take this long! After I get the data base figured out, I will connect
everything with PHP, or so the plan goes. The lack of resources has been a great
innovation driver to tremendously reduce complexity, but we had a little too
much lack 🙂 I did not have a good enough implementation plan – rather I did
not have specific time lines that if not met would let me know I needed to do
something radically different.
9/27/00 – Back to school
The meeting failed so I am going back to school. Taking an Oracle data base
class and looking for a job where I can hone my UNIX skills. I was only willing
to pay a little and a little is what I got
11/3/00 – Choose again (from Dan Simmons Hyperion trillogy)
We, my new partners and I, are actively looking for venture capital. They had
a great idea (selling Slashdot moderator points) and the more I thought about it the more it seemed really cool. That
idea is now called Castpoints.com based on the book Earthweb by Marc
Stiegler. The besieged earthlings use a framework based on true free market
economics to illuminate the best ideas from the cacaphony. People have to put
their money where their mouth is. The framework is a futures market (a zero
sum game with a specific ending time and a specific outcome – aka gambling,
speculating). To implement that is problematic in the USA due to high amounts
of regulation, paperwork, etc. At lease 2 groups are working on this anyway.
But what we came up with is just having people rate a post based on accuracy
and content, and seperately, if they like the viewpoint of the post. Raters
back this up with money so that they pay attention. Good posters are paid weekly.
This might turn out to be a cool and uesful thing.
In making the front end for Castpoints, we were again confronted with how to
organize, search and present hopefully huge amounts of data. We settled on structured
keywords. We will see how it goes??
11/14/00 – Addition of instantly buying stuff – We had this feature
long ago, but took it out in a simplification effort. But due to the focus on
Napster we put the function back in. It adds a lot of coding etc. but we think
it is worth it. The idea is, especially related to downloading music, that if
it is convenient, people will pay – certainly more then they do now. The security
is pretty basic but the vendor can do a few simple things that go a long way
to limiting one per customer. This Salon article
seems to bolster our viewpoint.
1/4/01 – It has been a while – We have been working on Castpoints and
think we have an excellent programmer. Hopefully the site will be up shortly.
When it starts making money 🙂 we will get the ball rolling with Infofitz. We
sent out another batch of letters to venture capitalists begging for money.
This time The tone of the letter had some hype. Interestingly enough, there
was less response to this letter – but it could be they had our names in a data
base and just……