Is AI a distraction???

AutomationI was recently in an exchange with a respected industry analyst where they stated that AI is not living up to its hype – they called AI ‘incremental’ and a ‘distraction’. This caught me a bit my surprise, since my view is that there are more capabilities and approaches available for AI practitioners than ever before. It may be the business and tech decision makers approach that is at fault.

It got me thinking about the differences in ‘small’ AI efforts vs. Enterprise AI efforts. Small AI are those innovative, quick efforts that can prove a point and deliver value and understanding in the near term. Big AI (and automation efforts) are those that are associated with ERP and other enterprise systems that take years to implement. These are likely the kinds of efforts that the analyst was involved with.

Many of the newer approaches enable the use of the abundance of capabilities available to mine the value out of the existing data that lies fallow in most organizations. These technologies can be tried out and applied in well defined, short sprints whose success criteria can be well-defined. If along the way, the answers were not quite what was expected, adjustments can be made, assumptions changed, and value can still be generated. The key is going into these projects with expectations but still flexible enough to change based on what is known rather than just supposition.

These approaches can be implemented across the range of business processes (e.g., budgeting, billing, support) as well as information sources (IoT, existing ERP or CRM). They can automate the mundane and free up high-value personnel to focus on generating even greater value and better service. Many times, these focused issues can be unique to an organization or industry and provide immediate return. This is not the generally not the focus of Enterprise IT solutions.

This may be the reason some senior IT leaders are disillusioned with the progress of AI in their enterprise. The smaller, high-value project’s contributions are round off error to their scope. They are looking for the big hit and by its very nature will be a compromise, if not a value to really move the ball in any definitive way – everyone who is deploying the same enterprise solution, will have access to the same tools…

My advice to those leaders disenchanted with the return from AI is to shift their focus. Get a small team out there experimenting with ‘the possible’. Give them clear problems (and expectations) but allow them the flexibility to bring in some new tools and approaches. Make them show progress but be flexible enough to understand that if their results point in a different direction, to shift expectations based on facts and results. There is the possibility of fundamentally different levels of costs and value generation.  

The keys are:

1)      Think about the large problems but act on those that can be validated and addressed quickly – invest in the small wins

2)      Have expectations that can be quantified and focus on value – Projects are not a ‘science fair’ or a strategic campaign just a part of the business

3)      Be flexible and adjust as insight is developed – just because you want the answer to be ‘yes’ doesn’t mean it will be, but any answer is valuable when compared to a guess

Sure, this approach may be ‘incremental’ (to start) but it should make up for that with momentum and results. If the approach is based on expectations, value generation and is done right, it should never be a ‘distraction’.

Advertisements

Is automation forcing divergent paths of quality vs. cost?

robots-too-humanI saw an interesting post: When Robot Writers Take Over, Will Freelancers Be Obsolete? The article was focused on freelance writing, but it did make me wonder about the whole concept of freelancing, in general.

The relatively fixed and easy to automate positions in many fields are ripe for automation. Those that require creativity or unique insight should be safe for a long time to come. In fact, automation could make the freelancers life less mundane and more interesting. It reminded me of a situation earlier in my career…

Back in the early 90s, I worked in the AI space for Electronic Data Systems (EDS). We focused primarily on solving problems for GM and the US government. Somewhere around here I have a coffee cup with the moto of the group: “Make it Work, Make it Real”. Unfortunately, the folks working in the group had felt it really meant that if we could make it work, it wasn’t really AI — since someone would always say that it was just regular old programming, no matter what innovative technique or esoteric language we used.

One of the projects I led was called Knowledge-based Tool Design. We were trying to automate tooling design for clamping and welding car parts using CAD techniques, a project far ahead of its time. Programmatically determining the right type of clamp and the correct way to swing it into place was too difficult spatially, for the time. We just didn’t have the compute power and the algorithms determine orientation and approach. A good human tool designer could see the solution intuitively.

We did figure out that people are not good at pulling together the bill-of-materials to ensure that the clamp and all the hydraulic and mounting components… were defined. We shifted our attention to defining that type of detail using computers — reducing the errors and rework later in the process.

Similarly, in other industries, there are so many annoying and resource intensive, low hanging fruit to be picked that the return on investment for tackling truly intuitive problems just isn’t there. That can all change though as better algorithms and computing capabilities develop.

There are a couple of ways this could go:

  • The intuitive functions will likely become more of a freelance function, since companies will not need (or be willing to pay) for those expert roles all the time and the work will be interesting.
  • The focus shifts to less high-quality designs that can be automated.

In any case, employment as we know it will be changing.

Simplicity, the next big thing?

Complex processRecently, Dynatrace conducted a survey of CIOs on their top challenges. Of the top six, almost all deal with concerns about complexity. There is no doubt there are numerous technologies being injected in almost every industry from a range of vendors. Integration of this multivendor cacophony is ripe with security risks and misunderstanding – whether it is your network or IoT vendor environment.

Humans have a limited capacity to handle complexity before they throw up their hands and just let whatever happens wash over them. That fact is one of the reasons AI is being viewed as the savior for the future. Back in 2008, I wrote a blog post for HP that mentioned:

“the advent of AI could allow us to push aside a lot of the tasks that we sometimes don’t have the patience for, tasks that are too rigorous or too arduous.”

IT organizations needs to shift their focus back to making the business environment understandable, not just injecting more automation or data collection. Businesses need to take latency out of decision making and increase the level of understanding and confidence. A whole new kind of macro-level (enterprise) human interface design is required. Unfortunately, this market is likely a bit too nebulous to be targeted effectively today other than through vague terms like analytics…  But based on the survey results, large scale understanding (and then demand) appears to be dawning on leadership.

The ROI for efforts to simplify and encourage action, should be higher than just adding a new tool to the portfolio ablaze in most organizations. We’ll see where the monies go though, since that ROI is likely to be difficult to prove when compared to the other shiny balls available.

Adding more complex triggers with IFTTT to control your home IoT

IoT HomeI have been using IFTTT for quite a while with my various IoT devices, doing simple things like turning on lights when my garage door opens or when there is motion near my home. One of the great things about IFTTT is its simplicity. If ‘this’ triggering event happens, do ‘that’. This simplicity is also one of the frustrating limitations of doing more complex tasks.

If I want to only turn on the porch lights when it is dark out, but keep them off when it is light, there is no way to do that natively within the system. They have thought about expansion capabilities through the ability to send and receive information from websites that can provide stateful information – IFTTT supports both trigger and response through a capability called webhooks. There are some free services to support this capability that you can experiment, if you don’t want to bring up your own website.

The one I looked at is: apilio.io and there is a good explanation of it on Medium. It is still in beta though.

Apilio has three building blocks:

  • Variables – to contain state information
  • Conditions – to evaluate variables
  • Logicblocks – to determine the actions from a Boolean analysis of the conditions

As an example, I decided to use Apilio with IFFF to turn my lights on when there was motion in front of my house but only when it is dark out. To accomplish this, I had to perform the following steps:

  1. Configure your IFTTT webhook connection on your webhooks settings/documentation page in IFTTT and enter that key into your Apilio profile. This key enables a secure connection between the website and IFTTT.
  2. Next define the variables in Apilio where you would like to store state information. In my case, I created a Boolean variable darkOutside. Note that there are URLs that when accessed will set the variable to True or False. You’ll need these URLs in the next step.
  3. Define two IFTTT rule Applets that set the variable to True at sunset and False at sunrise, using the capabilities of the Weather Underground trigger supported by IFTTT.
  4. Define a condition in Apilio that if darkOutside is True then it returns true, otherwise False. I named it DarkOutsideCondition
  5. Next, I made an Apilio logicblock that performs a simple logical AND operation that if it is triggered and the DarkOutsideCondition is True, it triggers applets back in IFTTT. I called the logicblock lights_on_when_dark. It has a URL to force its evaluation (from the logicblock show command) and a place to store a IFTTT trigger event name, if it is evaluated. One for True (called LightsOn) and another for when it is evaluated False (called LightsOff).
  6. Since I have a Ring doorbell with motion sensing that interfaces with IFTTT, I made an IFTTT Applet for when motion is sensed to initiate the evaluation trigger of the logicblock lights_on_when_dark.
  7. Now I just needed to create 2 more Applets, one to turn on the lights if it gets a webhook event called LightsOn and another if IFTTT gets the webhook event LightsOff that turn the lights on accordingly.

This may seem a complex but really only consists of:

  • a variable, a condition and a logicblock in Apillio
  • five applets in IFTTT:
    • set the darkOutside variable to true at sunset
    • set the darkOutside variable to false at sunset
    • receive a LightsOff trigger and turn the lights off
    • receive a LightsOn trigger and turn the lights on
    • catch the Ring motion sensor trigger and force and evaluation of the lights_on_when_dark logic block.

This is a fairly simple example. There are also some additional examples on the Apilio site, but hopefully this walkthrough will help you get started with enough context to overcome some of the areas that confused me.

I also have a IFTTT applet to turn the lights off at sunrise, just so there is another way to turn the lights off. I should have the ability to add some delay sensing so that I can turn the lights off after they have been on for a defined period (say a half hour), to keep my energy costs down.

IoT starts to come home

Over the years, I’ve played around with a few IoT solutions. Sunday, I decided to seriously tackle some outdoor lighting, by replacing one of my light switches with one that can be controlled from the Internet.

372.jpgI looked at a number of solutions and found that there are surprisingly few that will replace a 3-way switch (in fact the only 3-way switch I found was from GE and then I would need a controller…). After looking at my requirements, it appeared I only needed normal light switches and the one I chose was the Belkin Wemo® Light Switch. Fortunately, my house was already wired with switches that looked fairly similar, so my wife was happy with the result.

Thanks to Amazon same day delivery, the new switch was at my house at 6PM on Sunday.

I had it unboxed, wired in and controlled by IFTTT in under an hour. So now I can control when the light go on and off automatically and can turn them on manually from my phone. Not bad… for an hours work.

I have a few more projects that I am planning to do around the house, so I’ll write about those too, as well as anything I learn along the way. An example setting up a dedicated wireless LAN just for the IoT devices (to localize any security issues).

Action as a services moving forward…

action 002Had to laugh when I saw this post from Forrester titled: “Big Data” Has Lost Its Zing – Businesses Want Insight And Action since this is a song I’ve been singing for a number of years now. The first post about this I could identify was back in 2007.

Big data efforts should be measured in time-to-action, not time-to-insight. IT organizations need to be defining, developing and deploying systems of action.

To quote myself: “the organizations that can understand “normal” and focus the people on the areas that need their creativity will shine in the end. This relationship between situational awareness and automation needs to be part of organizational strategic planning, much more than what most architectural processes allow.”

Glad to see that analysts are recognizing this.

General fact checking falls to automation too?

true or falseFact checking by humans cannot keep up with the tremendous volume of information generated online. Now, computers can do fact-checking for any body of knowledge, according to Indiana University scientists, writing in a paper titled: Computational Fact Checking from Knowledge Networks. Computational fact checking may significantly enhance the ability to evaluate the validity of all the dubious information, the Internet is known for.

Using factual information from summary info boxes from Wikipedia as a source, they built a “knowledge graph” with 3 million concepts and 23 million links between them. A link between two concepts in the graph can be read as a simple factual statement, such as “Socrates is a person” or “Paris is the capital of France.”

The researchers aim to conduct additional experiments using knowledge graphs built from other sources of human knowledge, such as Freebase (the open-knowledge base built by Google).