Earn $4 Billion Dollars In Your Spare Time
A quick glance at a Google search (let your enemy supply you... I also use FEDERAL RESERVE NOTES!!) tells me the software development industry is worth $400+ billion per year.
Let's say I'm either inventive or know people who are. We have very good reason to believe we can save 10% of those costs somehow. With new technology, we expect to.
Suppose I want to dominate this market. But I'll have to invest 90% of my share to get there. Non-voting, obviously.
Our share of the market will be $40 billion per year. My share of this will be $4 billion.
Hey. It's good to be me.
But we'll assume I've got the vision, the self-discipline, connections, brilliance, or whatever, and a way of carving out a niche my competitors can't touch.
My share is then going to be 10% of a company that saves 10% of the costs of a job that costs $400+ billion per year, which is $4 billion per year.
I just have to make sure I save the marketplace $40 billion, no matter what it takes.
I can't, for example, complain that the hours are too long, the commute is too far, or any of that nonsense. I have to commit.
Well, most people aren't rich and they're not willing to commit.
And so they don't bother looking for ways to save 10% of the annual costs of creating software.
I look for ways of saving 90 to 99% percent of the costs of developing software while relatively quickly increasing their performance by a factor of 100,000, or else I'm not happy.
The industry is so chock-full of inefficiency for the sake of the coder's convenience. And that's fine. Their job is to create features, because investors want to please customers and users who demand features.
I'm writing this blog post with software that has a great feature I wanted, so I switched.The new software is buggy and ate half my last post, and so did the other software I used to save the previous post, but the features won out so I keep using it.
I may not save the other piece of software I trusted to save my post. I'll be looking for something that can handle a Fair-Use-length blog post or at least alerts me that my cut-and-paste has been truncated.
To make software development, as an industry, more efficient and more powerful, you need to continue to automate the manual labor. Coders know this and do this.
Then each coder's contribution (each problem-solver's work) becomes much more valuable.
They're using new tools that destroy less of their work at the last minute, for example.
In the future, we're going to use a new way of optimizing code. Currently, all problems are solved by stacking layers and layers of abstraction, letting people write high-level code for cross-platform compabitility.
Doesn't work very well, but it works. You sacrifice performance. Lots of performance. Billion dollar data center's worth. But it works.
That's because it's impossible and undesirable to have a billion assembly-language programmers re-writing for every software package for every hardware architecture ever invented.
But what if you could?
Maybe you want something like Ubuntu that works on the Rasperry Pi and works identically on the Apple Watch, that somehow both support a streaming, fully-playable version of Star Citizen.
Today, this would probably take a team of talented geniuses of indeterminate size, with no guarantee that anyone else but you wants to do those things.
You kinda want to, but you don't wanna pay for it. Can't it show an ad instead... and then automatically block that ad without asking it to?
I want automatically ad-blocking, ad-supported freeware that does everything I can ever imagine, even if I'm the only one who wants it.
And it's too slow on my Commodore 64. Yuck.
And I also want to be able to instantly make stuff without any talent or ability.
Oh, and I want it now. And all the time. And a new one every week. Or else I'll just take my business elsewhere.
Do we see the problem, here?
Infinite wants. Limited ability to deliver.
People on welfare who are entitle to our money, medical infrastructure, and Obama phones are putting an infinite tax on the creators, forcing them to deliver ever more pointless diversions to the most vapid, narcissistic, entitled, horrible people on the planet.
So it's time to offload some of that work to someone who gives a damn, don't you think?
Why don't we take the vapid, pointless, narcissistic, talentless people who want to create soul-sucking junk and foist it on their empty-headed friends, and make software development so easy and automatic that they're completely unaware it's what they're doing.
With "dumb" AI, this is becoming more and more of a reality. FaceApp does most of the work for you.
But when it comes to matters of taste or personal preference, you can artistically warp and ugli-fy or pretty-fy all your selfies.
The only part that's hard for computers to get right.
In the background, behind the app-makers, a team of hotshot compiler devs are figuring out how to optimize things for their new hardware platforms so apps seamlessly work cross-platform, and tools are created by big companies to make application portability virtually automatic.
Or as close to it as possible.
Apple, for example, wants to make it really easy for key companies to decide to support Apple products. Same with Samsung and everyone else.
At some point, we're going to have to admit that humans just aren't that good at developing software, even with all our marvelous capacity for reason, logic, and greebling together things.
And that all of us, including all the top tech companies are vastly better off, in the end, at finally letting computers do what they're best at.
It's time to start delegating, even when it's a "bad idea".
Because it will always be a bad idea, right up until the moment when it's the best idea ever.
Google isn't fast because the developers know how it works. It's fast because it's rewarded for being fast.
And this is what has made it the best for decades.
It's fast because it's efficient enough to justify throwing the best hardware at it forever.
It's efficient because nobody knows how it works. They didn't say, "I want to know how it works." They said, "I want it to be fast."
And fast was better than knowing why it's fast.
They also wanted it to be more accurate. But it's only accurate because we don't know how it works. Nobody knows.
If you've got software that works a hundred thousand times faster than another implementation, then $10,000 worth of hardware and electricity is doing as much work as a billion dollars of hardware and electricity.
You can't compete with it if you would rather know how it works.
I would rather be competitive.
But egos get in the way, especially when you can strategically stoke those egos. You say, "Here is the man who invented x, y, z. Don't you want to be able to say you're the man who did it?"
Well, no. I don't want to be able to say I did it. I want fast code.
If a million monkeys at a typewriters can do a better job, that's who I'll hire. Data centers are like very hungry ants or mice who devour electricity, but do the job.
I'd rather have the mice that work and eat like a pig than the men that don't work, no matter what you pay them.
Computers work for cheap. They do certain things extremely well, especially if you let them.
Right now, and for the last 30 years or so, computers have been very good at brute-forcing problems, like when they used brute-force computing power to beat Kasparov at chess.
But they're also good at something else. Something very different. At evolving algorithms so that brute force isn't necessary.
The power of evolution, especially in the digital realm, is that it quickly enough becomes more powerful than all the brute force computation on earth.
It's those who've recognized this fact who are getting richer and looting the world while humanity sleeps.
Because sleeping while someone takes our lunch is what the rest of humanity is good at.
The new paradigm of computing, the more powerful paradigm showed up, announced itself, and said, "Hey. I'm right here. I'm better than everything else and totally independent from your reliance on human intelligence."
Genetic algorithms have been better than humans for a long time, just as hand-held calculators have been better than humans for a long time.
In the realm of rational calculation, computers, our modern calculators, are billions of times faster than we are.
Those who paid attention to this fact now run the world to a greater extent than the rest.
Most of us rely on human intelligence to do what humans are incredibly bad at doing.
But if human beings were any good at making rational decisions, would they need computers?
We're only good enough at making decisions to recognize we're better off having their services than not.
But we're not good enough at making decisions to realize we're better off letting computers write all our most crucial, high-performance code for us.
The kind of code that's so efficient that noobdy knows how it works. That's so efficient, you'd gladly throw a billion dollars worth of hardware at it so that developers have jobs arranging their whole lives around a trillion dollar bit of code a computer wrote twenty years ago.
Let's say I'm either inventive or know people who are. We have very good reason to believe we can save 10% of those costs somehow. With new technology, we expect to.
Suppose I want to dominate this market. But I'll have to invest 90% of my share to get there. Non-voting, obviously.
Our share of the market will be $40 billion per year. My share of this will be $4 billion.
Hey. It's good to be me.
But we'll assume I've got the vision, the self-discipline, connections, brilliance, or whatever, and a way of carving out a niche my competitors can't touch.
My share is then going to be 10% of a company that saves 10% of the costs of a job that costs $400+ billion per year, which is $4 billion per year.
I just have to make sure I save the marketplace $40 billion, no matter what it takes.
I can't, for example, complain that the hours are too long, the commute is too far, or any of that nonsense. I have to commit.
Well, most people aren't rich and they're not willing to commit.
And so they don't bother looking for ways to save 10% of the annual costs of creating software.
I look for ways of saving 90 to 99% percent of the costs of developing software while relatively quickly increasing their performance by a factor of 100,000, or else I'm not happy.
The industry is so chock-full of inefficiency for the sake of the coder's convenience. And that's fine. Their job is to create features, because investors want to please customers and users who demand features.
I'm writing this blog post with software that has a great feature I wanted, so I switched.The new software is buggy and ate half my last post, and so did the other software I used to save the previous post, but the features won out so I keep using it.
I may not save the other piece of software I trusted to save my post. I'll be looking for something that can handle a Fair-Use-length blog post or at least alerts me that my cut-and-paste has been truncated.
To make software development, as an industry, more efficient and more powerful, you need to continue to automate the manual labor. Coders know this and do this.
Then each coder's contribution (each problem-solver's work) becomes much more valuable.
They're using new tools that destroy less of their work at the last minute, for example.
In the future, we're going to use a new way of optimizing code. Currently, all problems are solved by stacking layers and layers of abstraction, letting people write high-level code for cross-platform compabitility.
Doesn't work very well, but it works. You sacrifice performance. Lots of performance. Billion dollar data center's worth. But it works.
That's because it's impossible and undesirable to have a billion assembly-language programmers re-writing for every software package for every hardware architecture ever invented.
But what if you could?
Maybe you want something like Ubuntu that works on the Rasperry Pi and works identically on the Apple Watch, that somehow both support a streaming, fully-playable version of Star Citizen.
Today, this would probably take a team of talented geniuses of indeterminate size, with no guarantee that anyone else but you wants to do those things.
You kinda want to, but you don't wanna pay for it. Can't it show an ad instead... and then automatically block that ad without asking it to?
I want automatically ad-blocking, ad-supported freeware that does everything I can ever imagine, even if I'm the only one who wants it.
And it's too slow on my Commodore 64. Yuck.
And I also want to be able to instantly make stuff without any talent or ability.
Oh, and I want it now. And all the time. And a new one every week. Or else I'll just take my business elsewhere.
Do we see the problem, here?
Infinite wants. Limited ability to deliver.
People on welfare who are entitle to our money, medical infrastructure, and Obama phones are putting an infinite tax on the creators, forcing them to deliver ever more pointless diversions to the most vapid, narcissistic, entitled, horrible people on the planet.
So it's time to offload some of that work to someone who gives a damn, don't you think?
Why don't we take the vapid, pointless, narcissistic, talentless people who want to create soul-sucking junk and foist it on their empty-headed friends, and make software development so easy and automatic that they're completely unaware it's what they're doing.
With "dumb" AI, this is becoming more and more of a reality. FaceApp does most of the work for you.
But when it comes to matters of taste or personal preference, you can artistically warp and ugli-fy or pretty-fy all your selfies.
The only part that's hard for computers to get right.
In the background, behind the app-makers, a team of hotshot compiler devs are figuring out how to optimize things for their new hardware platforms so apps seamlessly work cross-platform, and tools are created by big companies to make application portability virtually automatic.
Or as close to it as possible.
Apple, for example, wants to make it really easy for key companies to decide to support Apple products. Same with Samsung and everyone else.
At some point, we're going to have to admit that humans just aren't that good at developing software, even with all our marvelous capacity for reason, logic, and greebling together things.
And that all of us, including all the top tech companies are vastly better off, in the end, at finally letting computers do what they're best at.
It's time to start delegating, even when it's a "bad idea".
Because it will always be a bad idea, right up until the moment when it's the best idea ever.
Google isn't fast because the developers know how it works. It's fast because it's rewarded for being fast.
And this is what has made it the best for decades.
It's fast because it's efficient enough to justify throwing the best hardware at it forever.
It's efficient because nobody knows how it works. They didn't say, "I want to know how it works." They said, "I want it to be fast."
And fast was better than knowing why it's fast.
They also wanted it to be more accurate. But it's only accurate because we don't know how it works. Nobody knows.
If you've got software that works a hundred thousand times faster than another implementation, then $10,000 worth of hardware and electricity is doing as much work as a billion dollars of hardware and electricity.
You can't compete with it if you would rather know how it works.
I would rather be competitive.
But egos get in the way, especially when you can strategically stoke those egos. You say, "Here is the man who invented x, y, z. Don't you want to be able to say you're the man who did it?"
Well, no. I don't want to be able to say I did it. I want fast code.
If a million monkeys at a typewriters can do a better job, that's who I'll hire. Data centers are like very hungry ants or mice who devour electricity, but do the job.
I'd rather have the mice that work and eat like a pig than the men that don't work, no matter what you pay them.
Computers work for cheap. They do certain things extremely well, especially if you let them.
Right now, and for the last 30 years or so, computers have been very good at brute-forcing problems, like when they used brute-force computing power to beat Kasparov at chess.
But they're also good at something else. Something very different. At evolving algorithms so that brute force isn't necessary.
The power of evolution, especially in the digital realm, is that it quickly enough becomes more powerful than all the brute force computation on earth.
It's those who've recognized this fact who are getting richer and looting the world while humanity sleeps.
Because sleeping while someone takes our lunch is what the rest of humanity is good at.
The new paradigm of computing, the more powerful paradigm showed up, announced itself, and said, "Hey. I'm right here. I'm better than everything else and totally independent from your reliance on human intelligence."
Genetic algorithms have been better than humans for a long time, just as hand-held calculators have been better than humans for a long time.
In the realm of rational calculation, computers, our modern calculators, are billions of times faster than we are.
Those who paid attention to this fact now run the world to a greater extent than the rest.
Most of us rely on human intelligence to do what humans are incredibly bad at doing.
But if human beings were any good at making rational decisions, would they need computers?
We're only good enough at making decisions to recognize we're better off having their services than not.
But we're not good enough at making decisions to realize we're better off letting computers write all our most crucial, high-performance code for us.
The kind of code that's so efficient that noobdy knows how it works. That's so efficient, you'd gladly throw a billion dollars worth of hardware at it so that developers have jobs arranging their whole lives around a trillion dollar bit of code a computer wrote twenty years ago.
Comments
Post a Comment