7/27/2016

Intelligent Machines will need 5 Basic Senses

You can train a model by feeding tons of data and tuning each of the layers to decipher tiny dots into images into things into people and such through pattern recognition machine learning tools.

You can do the same with audio.  In fact, you can take some spoken word, interpret what was said and then send it back out in another language.  One input, translator, many output possibilities.

Likewise, you can receive many inputs, translate, send them all out in one specific language.

So we can process sound and sight, apply machine learning techniques and understand the content of the media.

What about smell.  Do we have any device that can interpret smells.  For instance, you've got a bag of popcorn in the microwave for 30 minutes, strong odor emitting, can a sensor pick up the burning smell and trigger an alter through machine learning?

When we capture sound or visual media, can we also pick up the surrounding environment such as odor.  Create a machine learning model that is trained on smells?

How about taste?  Can a machine tell the difference between Coke and Pepsi?

How about touch?  Difference between a dolphin's skin and a porcupines skin?

I believe those are the 5 basic sense we humans possess?

Some say we have a 6th sense, intuition.  Fairly certain we don't have a model for that one.

How about taking a scan of an object, and determining the molecules that make up the object?  Based on the arrangement, identify the object based on the model?

How many levels of attributes can we capture simultaneously?  Sight, sound, touch smell and taste.  How about location, who's in the room with you, time of day, etc..

It seems if we are to build intelligent machines that interact on daily basis, they should at least have the 5 basic senses.  Otherwise, an inherent disadvantage.

What if we could scan a portion of space, like a 2 x 2 x 2 location, identify each molecule contained within that space, label each, and track their movement over time.  What if we did this with every seeable molecule inn the Universe.  Store in big data, play scenario back and forth, then predict future behavior of each molecule.  Suppose we'd need more processing power, enter Quantum Computers stage left.

Having the ability to process more than 1 logical bit decision per millisecond, it can process multiple depending on the number of Qubits, you could exponentially apply probability calculations on possible movement of molecules over time.  And like the monkey typing on a keyboard, eventually write complete work of Shakespeare, maybe we could get better at predicting movement of things over time, for starters, the weather.  Planets.  Galaxies.  Electrons.  Quarks.  You name it.

7/22/2016

Multiple Barriers for Entry into Advanced Sciences

The scientific model is used to validate things as "facts".  We state a hypothesis, runs tasks over time, documenting minute details during the testing, and if the outcome matches our guess, it's fact.  If it fails, then its incorrect and adjustments are made to the hypothesis and tried again.

Some things can't be measured, duplicated or the costs to run experiments is too high.  Or we need to build a billion dollar infrastructure to test our theories, which takes years of effort and hoops of red tape to get funding, if you're lucky.

And of course, there's the inherent mathematics behind everything to prove things.  Sometimes, new math needs to be created.  Lots of discoveries happen when some person goes through the stacks at on old library, finds a gem in an old book and re-discoveries something that a non-famous person discovered hundreds of years ago.

And many of our presumed facts are clearly inaccurate or just wrong.  We sure have come a long way over the years.  Except somehow they were able to build monoliths of magnificent proportion thousands of years ago before advanced math and track the stars in the sky before telescopes.

Lots of discoveries happen over the years by people on the sidelines, tinkering with advanced subject, who don't really have a stake in the game.  The thing is, in order to advance mankind with new ideas and inventions, we need to have critical thinking skills along with freedom to roam and explore new territory.  When you're embedded in the one of the sectors of a specific science, and you derive some new theory, and it doesn't align with established thinking, you're labeled a crackpot, risk losing funding and reputation.

So there's a mote across basic science preventing non-professionals in, lots of fierce competition among the intellectual elite and the entry takes years of schooling and advanced degrees that take 25 years.  With so much invested, the pecking order is well established and the experts dictate 'truth' downhill.

There's a history of brilliant minds who never got credit for their amazing ideas.  The scientific model, which states you must prove something in a lab which is documented and reproducible, keeps a stranglehold on new discoveries.  As does the required funding, authorization, education requirements, peer support, testing infrastructure and access to material.

We need to come out of the dark ages and provide new entry ways to some of the advanced concepts and technologies to propel mankind forward.  Our legacy system needs an upgrade.

Data Vendor Sprawl

We talk about the data warehouse back in the day, IT had too much control, users weren't getting data behind closed walls.

Enter self service Business Intelligence, stage right.  Whip out a credit card, get sweet talked by a new vendor, bam, you've got yourself a new Business Intelligence infrastructure.

Guess what, so did the department to your left, and your right.  Now your IT team does an assessment, we've got 5 different vendor Business Intelligence platforms to contend with.

Support, finding qualified developers, training staff, backups, data mismatches, reports don't match.

Freedom comes with a cost.  The cost is fragmented technologies and data vendor sprawl.

We all want freedom.  To feel empowered.  To make a difference.  In order to have that freedom, we need structure.  Otherwise its a free for all.

Sure we don't understand much of what the IT folks are talking about, about Star Wars and teleportation and some new video game they played over the weekend for 54 hours straight.  But to democratize data, we need someone to steer the ship.  I say the Chief Data Officer is designed to handle such tasks.

It seems the CDO role is increasing in larger organizations, which is great.  Data platforms need to be consistent, to reduce costs, get bundled savings, similar developer skills, etc.

Because if you have 5 different vendors under one roof, and all 5 release new version, and 5 developers leave, the business doesn't want any hiccups in anything, deliver or perish.

To alleviate data vendor sprawl, you need some one at the helm, either a Chief Data Officer or Business Intelligence Competency Center or part of the CIO role, someone needs to be responsible and implement some corporate structure.

7/15/2016

Top 14 Skills for Any New Client Engagement

When starting a new client engagement, you have to learn many things.

  1. People.  
  2. Hierarchy.  
  3. Processes.  
  4. Procedures.  
  5. Domain Knowledge.  
  6. Technology.  
  7. Permissions.  
  8. Passwords.  
  9. Change Management.
  10. Listen.
  11. Learn.
  12. Provide Value.
  13. Find the biggest pain point
  14. Then create working solution.

Always exciting to start new projects.

7/12/2016

The Newest Most Challenging Technology Today - Quantum Computing

If you like intellectual challenges, you have a few options to sink your teeth into.

First, there's the elusive phenomenon called Artificial Intelligence.  There's the "weak" type, like Cortana or Siri, which exists today in commercial applications.  And there's the "strong" called Artificial General Intelligence, which is still pending, time to delivery 65 years and counting.

And still yet, an even tougher challenge exists.  Called Quantum Computing.  Going past the traditional microchips designed on the binary algorithm of dual states, zeros and ones, the new concept tries to leverage the quirky physics called "Quantum Physics".

Such that the state can be zero or one, or zero and one, or partial zero and partially one.  The unit of measurement is called "Qubit".  This allows more combinations simultaneously, thus, more processing power in shorter time.

The Qubit requires an unusual environment, very cold temperatures and isolated from outside interference.  The machines are costly to build, maintain as are the new scientist that work on them.

In essence, they are building the new equivalent of the transistor.  The competition to build a working quantum computer is intense.  Many high tech organizations with deep pockets are funding these decade long projects.  With hopes to build a working machine and the blueprint for the next wave of inventive architectures and solutions.

Here's a link that describes Microsoft's vision and status thus far:

https://www.technologyreview.com/s/531606/microsofts-quantum-mechanics/

And a YouTube video of one of Microsoft's top Quantum Scientist:


This stuff is fascinating.  I don't understand any of it.  Well, maybe some.  I get tangled in the de-coherence and my superposition of vector state is entangled in sub-particles.  Thus, not sure if I understand completely, not at all, both, or somewhere in between.

Lastly, here's a link on a new language designed to process Quantum algorithms on Quantum Machines with a catchy name "LIQUI>" (liquid):

https://www.microsoft.com/en-us/research/project/language-integrated-quantum-operations-liqui/

If you understand this stuff, have a PhD and a good understanding of Physics, Machine Learning and Computers, there's definitely a job out there waiting for you.

Thanks for reading~!

7/07/2016

Artificial Intelligence Touches Every Aspect of Human Existence

There is something very profound taking place today.  And that is Artificial Intelligence.  And the reason is AI is "the" disrupter like no other.

Why is that?

It cuts through every segment of life.

It starts with data.  It takes data to train a Neural Net model.  Somebody gets to choose what data to use.  Typically, the people sponsoring the project get to choose.  And because it's their dollar, they get to decide the intent.  Should AI be used for profit, to better mankind, or for control and power.

After 65 years of false starts, there's finally some progress in the field.  And the consensus at the moment is traditional programming since the 1950's will not solve the problem.  The closest theory to attain maximum benefit is to mimic the human brain.  The machine like part that stores memories, retrieves thoughts, learns over times, through the associate of patterns to things and events and outcomes and benefits.

The brain builds a simulated model of the world.  Your model is different than mine.  Because it's learned through personal experience over time.  Our interpretation of the world is determined by our input sensor, eyes, ears, nose, taste, touch and intuition.  However, each sense gets interpreted by the brain using the same algorithm.  In actuality there's no difference between hearing and seeing.  The brain matches patterns, to reinforce it's model, which has been feed data since birth.

AI brings up philosophical debates.  Is it moral and ethical for robots to kill a human or another robot.  Can robots be used to carry out the dirty work of humans with no ramifications to the person issuing the orders.  Do robots have rights or are they property.

AI brings up religious issues.  Does a human and/or machine have a soul.  Does the soul die at death or go somewhere else.  Is there a higher power and are humans somehow attached metaphysically.

AI brings up economic issues.  If intelligent machines can perform some tasks better than humans for a fraction of cost, will companies toss humans aside to maximize profit and reduce costs.  This brings up the basic idea of the humans purpose on this planet.  If not perusing monetary gains through gainful employment what exactly is our purpose.  This exposes the overpopulation of the species and the dwindling resources to support x billion people.  This employment issues has ripple effects and is brushed over as someone else's problem.  As is the moral issues of intent.

AI brings up biological issues.  Are robots living creatures that can self replicate.  Can a person's soul be uploaded to a computer for everlasting life.  Will a hybrid version exist part human part machine android.  Can future babies be ordered from a catalog, picking features based on prices which only rich can afford.

AI brings up physics issues.  Can Quantum Physics be integrated into machines to leverage the bizarre behavior of the hidden laws of the Universe.

AI brings up our lack of understanding of consciousness.  What is it, how do we measure it, who and what has it.  A complete mystery with very little agreement from the top minds over thousands of years.

AI brings up our education system in that we need smart people to tackle these tough questions.  Yet education is getting so expensive and doesn't guarantee future employment, our population could be lacking the critical problem solving skills to solve tough problems.  And as automation increases, why would people attend advanced school at high costs if computers can outwork them with no paycheck.

AI brings up psychological questions.  If the goal is to mimic the human mind, are we going to limit it to just intelligence and leave out emotions like hatred, revenge, envy, spite, anger and even love.  Can a robot get depressed.

Clearly, there are lots of unanswered questions.  However, it's very apparent that AI has a lot more to do with non technological questions than the actual technical side.  And that's because it seeps into every nook and cranny of our human existence.  The problem must be attacked from multiple angles.  And we clearly don't have enough brain power to think through all these tough questions at the same time.

As we know, a democrat's model of the world is different from a republican's model.  A person sitting in jail's model is different from a politician's model.  Because each person has a unique perspective of the world based on unique experiences.  Sure, there's some overlap, but the model's explain our personalities, which are unique.  Our model's explain our consciousness.  Our model's explain this great vast Universe in a way that's easy to understand and promotes our survival.  Our model's get fed every minute of every day, new stimuli, that either learns something new by creating new memories and patterns in the brain, or reinforces existing beliefs so the synapses follow the same patterns they're used to.

If the goal is to create smart machines, to solve tough problems that humans have wrangled over the centuries, to find the hidden patterns in the Universe and explore the unexplored in pursuit of science, then it makes pretty good sense.  If the goal is to put the humans out to pasture from the workforce because they no longer provide value in supporting the economy because they are financially leveraged to the hilt with no earning power, that's not too good.  And if the goal is to create a robotic layer between the have's and have not, to police the population and become the new military force, that might not benefit the average Joe.

It comes down to who's financing the AI projects.  Because they're the ones making the decisions.  Because the average Joe has not earthly idea what's going on, and are not available to participate in key decisions.  The education research has basically left the Universities because the Corporate world has the deep pockets and can move much faster without layers of decision makers.  Innovation is taking place this very second.  And the impact could alter the course of the planet.

Lots to think about~!

7/06/2016

What is Consciousness ? - Three Stages of Consciousness | Michio Kaku

Good discussion here on consciousness with an actual definition with 3 levels and units of measure.

And the world is not predeterministic as Newton and Einstein formulated, due to Quantum Mechanics theory that contains uncertainty.

Lastly, the brain is a biological machine, with its core being a Neural Net, with Neurons firing Ions across synapses, that occasionally leak.  And criminal minds are to blame for the criminals actions, defective.

Says computers will eventually pass the Turning test and be slightly less than an actual human brain, very indistinguishable.

And people have been trying to mimic the brain using computers, which is incorrect, should be using Neural Nets that can learn and rewire themselves from time to time.

Have a listen~!


Dual Neural Nets Observing, Predicting and Learning from Each Other

I think it works like this.  Your mind watches the world.  Through input stimuli.  Including sight, sound, touch, and your unattended observation device called the third eye or intuition.

The brain, based on fight or flight, always aware of possible danger, instructs the body to perform actions, by sending signals to ligaments to move arms and legs.

As the action occurs, the observational aspect watches and determines next course of action, based on results of prior action.

Because of the overload of stimuli, the mind has the ability to condense information packets, by storing past events and how objects relate to things.  Basically like a fast retrieval mechanism where it doesn't have to process every bit of information, it sees things, determines if threat, then performs action, waits for feedback and then begins again.

So in teaching computers how to exist in modern world, it has to learn the same process.  In other words, how does it learn "common sense".

Some very smart people are working on possible solutions.  One I read about yesterday, the advanced AI team at Facebook, is combining dual neural networks to mimic our method of learning.

One Neural net will be given a task, and it will predict the outcome.  The other Neural net is aware of the event to happen, it knows what the other Neural net thinks will happen and then the action occurs.

The opposing net see the results, compares to what it think should have happened and updates its model appropriately.  It learns what should happen and what actually happened.

It seems like a good approach, sort of like mirrors.  I wonder what would happened if there were 10 Neural Nets watching the event.  Like a community learning session.  Like a deep learning observational network.

That way, you could throw in more variable instead of simple tasks.  Because a limited single domain expert can not survive in the world.  You need a network of Neural Nets with multiple domains integrated and working in real time, always observing, predicting, watching and updating.

Seems like a good hypothesis to test.  I don't do this stuff for a living, but its interesting to keep up with the latest developments in the hottest field going today.

And there you have it~!

7/04/2016

Thoughts on Artificial Intellience Trends

In the world of Artificial Intelligence there's lots of duplication across the postings out on the Internet.  Some say it will never happen, others say we have it now in high level form, while others say it should be here shortly.

AI is actually of subset of Cognitive Intelligence.  Siri and Cortana have been here for a while, while robotics is coming on strong, and the ability to process huge data with a deep learning layer sitting on top, allows for predictable outcomes, categorizations and classification.

Automation is mixed in there as well, with recent chat bot software available to the public.

Internet of things is still in the limelight although security concerns have probably limited its widespread adoption.  Along with different languages, protocols, networks, proprietary verses open source solutions.

I haven't kept up with Virtual Reality much or 3-D printing so I can't say for sure where that stands in the chaos.

Listening to a podcast earlier today, they discussed AI and two main concerns.  Will it take all the jobs and will it be evil?

Automation is coming, beginning with repetitive jobs that are mindless and don't deviate much in processes or education levels to perform basic tasks.  Robots and AI will automate those quicker than other jobs.  But the complex jobs are not completely safe either.  Medical advisors that can scan Radiology reports a lot quicker than humans, with less errors, more precision, faster and they don't fatigue.  Automation will trickle upwards into white collar jobs almost guaranteed.

How do people defend against it, they claim education.  That sort of directly opposes the current debates on whether a 4 year college degree is worth the time, money and effort as students graduate with tremendous debt, little chance of full time work in their field, a shrinking job market, baby boomers who refuse to retire and job simply disappearing.  Likewise, the quality of education today has been focused on "testing" and students being taught material "just' to pass the tests.  There's definitely an opportunity for the education system to leverage technology in real time using the internet to keep up with rapid change and deliver quality content across the board to ensure every child has equal chance regardless of other factors including location, parents income level and culture differences.

The other topic of interest was whether or not AI would be used for evil, in that the AI things would turn on their creators and dominate the world.  They mentioned a few times that it depends on their "training".  Who trains these AI beings using what data with what intent.  However, perhaps it boils down to who's funding the program and the people behind the program and what is their intent.  Even still, AI beings can learn, self teach and perhaps re-write their own operating systems and internal kernels such that they no longer take orders from anyone.  At this point, you can't say that could never happen.

All in all, these are exciting times in the world of Data, AI, Robotics, Automation, Deep Learning, IoT, Quantum Computing, Virtual Reality and 3-D printing.  Just wondering if this could be the last generation to go through the system with a chance of self sustaining work to support a family through gainful employment.

And there you have it~!

Here's the link from YouTube: Artificial intelligence will change everything | Ginni Rometty, CEO IBM | Code Conference 2016

7/01/2016

My New Online Course is now Available

If you're interested in learning about Azure Data Platform and Cortana Analytics, you can check out my new course at Experfy.com.

Here's the link: https://www.experfy.com/training/courses/gain-competitive-advantage-using-microsoft-azure-data-platform-and-cortana-analytics

It was a fun project to build the course outline, the demos and film and edit everything.  The people at Experfy are really nice to work with.

You can also check out my blog there:  https://www.experfy.com/blog/author/jonathan-bloom

Thanks~!

Get Sh#t Done!