Machine Learning Archives - NuTech Services LLC

Dec, 2021

Do Humans Create Bias in the AI We’ve Developed?

Science fiction shows artificial intelligence to be an entity compelled purely by logic, driven only by objective facts. AI tools used by businesses and in the real world, however, are a far cry from this perception. AI systems have some biases in their operations. Let’s take a look at some of them and how you can resolve these issues.

What Kind of Biases Have AI Systems Demonstrated?

There are several biases that AI can display. Here are some of them:

Sampling Bias: This occurs when AI is only given part of a population or a selection of samples rather than a purely random process.
- Voluntary Bias: voluntary bias specifically refers to how a population’s results are artificially skewed by their willingness to participate.
Design Bias: This bias is a flaw in the process itself which leads to flawed outcomes. In AI, the issue is most often found in the dataset.
Exclusion Bias: This type of bias occurs when specific data is intentionally removed or omitted, and it ultimately yields fewer or less valuable insights.
Label Bias: This bias occurs when the data is not labeled correctly. See below for the two types of label bias:
- Recall Bias: This form of bias appears in data that has been mislabeled and annotated inaccurately.
- Measurement Bias: This division of label bias is the result of inaccurately or inconsistently taken data points.
Confounding Bias: This bias happens when external variables are pulled into the equation or directly influence your data set, leading to inaccuracies in the final product.
Survivorship Bias: This type of bias occurs when only data that has made it through the selection process is considered. For instance, World War II researchers made this error when examining fighter jets to better reinforce them. By only examining jets that survived the trip back from a combat mission, the most useful information (where the planes that went down were hit) was ignored.
Time-Interval Bias: This bias occurs when data from only a specified period of time is analyzed rather than the complete set.
Omitted Variables Bias: This bias happens when data collected is cherry-picked and only certain variables are considered, thereby skewing the results.
Observer Bias: This is essentially confirmation bias, where an individual only considers data that matches their own values or goals rather than the complete set.
- Funding Bias: This variety of observer bias comes when the interests of a financial backer leads to the data being skewed.
Cause-Effect Bias: This is when correlation is mistaken for causation, or when two events happening at the same time are thought to be because of each other without taking into consideration other factors.
Model Over/Underfitting: This bias occurs when the analytical system, or model, can’t see the big picture or is not able to grasp patterns appropriately.
Data Leakage: This occurs when two sets of data that are to be compared share data, like when you are comparing a certain time period to your predictions.

Where Do These Biases Come From?

In most cases, these biases are formed from the system or, more specifically, the user of that system.

AI Bias is Just an Extension of Human Bias

Whether it is error based on prejudice or assumption, most biases can be traced back to the user. For example, let’s say that you want to determine the most important part of your services to your clients. In this oversimplified example, the algorithm powering the AI could be perfectly put together, yet the data used could muck up the results. For instance, if the data was specifically and exclusively collected from Facebook followers, then the accuracy of the data will be skewed in a certain way (sampling bias and voluntary bias, as your followers need to opt into providing you with this data).

This is but one example of AI being unable to perform its assigned tasks, so to prevent this from happening, you must approach the design of your AI systems with an awareness and willingness to avoid biases.

That’s right—it takes human awareness to help AI do its job in an appropriate manner.

How Can Bias Be Avoided in AI?

You can take certain steps to keep biases from impacting your AI systems. There needs to be a capability for a human being to observe the processes and catch its mistakes, as well as the opportunity to update the systems to accommodate any adjustments as needed. There must also be standards placed on the data collected to ensure that opportunities for bias are minimized.

Your team members will also have to remain aware of these biases while they are working with your data. These biases are generally sourced from human biases, meaning that they can influence your business even if you aren’t using an AI system. In other words, you need to make sure that your staff are both aware of and actively avoiding these biases when processing, collecting, and analyzing data.

What are your thoughts on AI and its uses in the business world? Be sure to leave your thoughts in the comments.

Jan, 2021

Artificial Intelligence and the Tools Designed to Improve Business

AI has been one of the leading innovative topics to hit the technology world over the past couple of years. At first, people thought that AI was only the process of creating machines that will inevitably destroy the human race, and didn’t understand that algorithmic machine learning could have stark benefits for business and society. Today, AI can be found in all types of different pieces of software. Let’s take a look at a couple of ways AI is currently being used in business.

#1 – Cybersecurity

One of the most important uses of AI is for cybersecurity, most of which is identifying actual threats and eliminating them before they can cause any problems for a business. There are a lot of potential threats out there and today, IT professionals are using AI to avoid spending time on situations that turn out to be non-issues. AI can be used to detect intrusions, identify vulnerabilities in software, and find malicious code that has already been installed on the system.

#2 – Customer Service

AI’s most noteworthy application has been the incorporation into customer relations. Chatbots and other technologies that are fueled by machine learning can provide a lot of value for most organizations that simply cannot afford to employ a complete product or service support team. Users may not even realize that they are engaging with an AI as many newer solutions learn rapidly to provide customers with a fast, reliable interface in which to get support.

#3 – Operational Efficiency

For some time, automation has been the name of the game when businesses attempt to streamline operations. Today, AI is beginning to provide a more diverse set of companies the opportunity to leverage more sophisticated tools. Since AI is constantly evolving and developing, more and more businesses are able to build tools around AI/machine learning than ever before. AI allows businesses to automate more of the mundane and repetitive tasks that have hindered productivity and progress, reducing costs, and providing a substantive boost in efficiency.

There may not be AI beings peppered among us yet, but AI is making a big push to be the most important emerging technology of the 21st century. If you would like to learn more about getting the results your business wants from its relationship with technology, return to our blog or give us a call at 810.230.9455 today.

Jan, 2019

The Rise of Machine Learning for Small Businesses

A.I. is one of those technologies that captivates the imagination with endless possibilities. You can’t turn your head these days without using something integrated with early artificial intelligence. Machine learning platforms, which are very rudimentary forms of A.I. are now being used to improve many of the tools a small business uses. Today, we will briefly go into what machine learning is and how small businesses are using it to their advantage.

Smarter Machines?
The first thing you have to understand about machine learning is that it is just a branch of A.I. As such, it basically describes a method of analyzing huge amounts of data in which prospective models of problem solving are automated. In layman’s terms, the computer solves basic problems without human intervention. This is significant for a few reasons, but the most glaring is that businesses today, and even small businesses, take in a lot of data. Traditionally, humans were paid to go through all this data, or it was archived and disregarded until there was a major question.

This shift is in the manner in which these machine learning systems analyze data, identify patterns, and make decisions from the analysis of those patterns.This seemingly advanced technology is being deployed at a dizzying rate and is beginning to surround us all, in our phones, on search engines, and in the systems that we manage our business data. For the small business, there are opportunities to utilize this technology to help carve out a larger market share.

Small Business Machine Learning
The philosophy behind utilizing machine learning for a business is just an extension of the overarching strategy of deploying automated systems to cut down on personnel and human resources costs. This strategy has worked in several sectors, albeit with automated systems that were more pre-programmed than “smart” like the machine learning systems.

An issue many small businesses will run into when deciding whether or not to try and innovate to the point where machine learning is an option, is where exactly it fits. Other questions persist as well. They include:

What is the capital and operational investment of deploying this technology?
How it will have to be utilized to provide a competitive advantage?
What systems can be improved through the use of machine learning?
How much time do you have until you would be at a competitive disadvantage if you didn’t invest in the technology?

Once you’ve ascertained how exactly to deploy machine learning, you can go ahead implementing it where you feel it is warranted.

Benefits of Machine Learning
A.I. and machine learning carry with it specific benefits. Some of them include:

Forecasting business – What might be the most popular use of machine learning, the technology can be used to replace data and service analysis while being used to replace processes that were done manually or not done at all.
Customer service – Chatbot technology essentially automates the customer service experience by directing customers to certain solutions based on query.
E-commerce – Rolling out machine learning for your e-commerce site helps customers by adapting to customer behavior. Giving prospects and customers easy access to obtain the the products and services they are most interested in is sure to improve revenue generation.
Reputation management – Machine learning can be set up to analyze internal and external data sources to monitor brand popularity. When it finds negative sentiment, you can use the information provided to smooth the situation over, fast.

These are only four of the many processes that can really help improve your business. As the technology improves, more A.I. and machine learning applications will be used to manage, maintain, and streamline a lot of your business’ slow-moving manual processes.

Machine Stupidity
These technologies are extremely advanced and are programmed to learn for themselves, so oversight of them may be difficult. If you are one of the many business owners that have a difficult time trusting technology, machine learning may not be for you. These constructs tend to be less predictable than typical analysis, because it is so thorough. Also, you may run into problems getting all of the data that needs to be analyzed into the system, providing results that may not be accurate for your entire business.

One could see how a small business owner can be in ‘wait and see” mode, but the problem with that strategy, however, is that your competitors may be using it now to improve their business. If you are interested in learning more about how this emerging technology can be used to make your business better, contact the professionals at NuTech Services today at 810.230.9455.

Jun, 2018

How Does the A.I. in Reality Measure Up to Hollywood’s?

Figuring out how to utilize platforms that depend on machine learning to boost an organization’s bottom line is one of the biggest puzzles for every modern business owner. After all, seemingly every new technology concept can be leveraged into enhanced profitability if it is rolled out right. In this case, many organizations have found ways to use human-created machines to learn how to do tasks that would be deemed too expensive if humans were to do them.

Over the years, A.I. has been a frequent topic of discussion, and one that fiction authors (especially those in science-fiction) have used for all types of stories. In Hollywood, the artificially intelligent character has been around for a long time, a lot longer than the A.I. businesses are using to enhance their profitability. Today, we are going to look at how A.I. is portrayed in media and how it differs from the reality of modern A.I.

The Start of A.I. in Reality
In 1956, 30-some scientists met at the Hanover Inn on the campus of Dartmouth College in New Hampshire to discuss a “strange new discipline”. The talks were about how to build a machine that could think and went on for weeks. What came to be known as the “Dartmouth workshop” founded a quest for A.I. The discipline almost died off several times, but if you look at the world we live in today, it’d be hard to consider that. These days it seems like every business is using some sort of software platform that features what those at Dartmouth a half-a-century ago could only dream about. Machine learning has seen major innovations in many different industries, and we are closer than ever to deep learning–the innovation needed in machine learning to create machines that think like we do.

The Start of A.I. in Hollywood
In Hollywood, however, deep learning is a thing of the past. Machine sentience is commonplace and stories of A.I. are typically approached as commentary about the tyranny and hubris of human beings. A.I. works for so many different types of story arcs as setting an A.I. up as the hero works, setting them up as the victim works, and setting them up as a villain works. In fact, since humans haven’t mastered the technology, writers do what they do best: use creative license to create A.I. characters that are more like humans than machines. The first robot A.I. was Robby the Robot in 1956’s Forbidden Planet, but 1931’s Frankenstein was the first time an artificial being was brought to life on the big screen. A.I. is often used as a plot device for entertainment’s sake, or as commentary, but media hasn’t been able to completely represent where we are at with the technology today because, thus far, A.I has been created to help humans solve problems, not to actually have artificial consciousness. Here are a few movies that represent different uses of A.I. and how they stack up against modern A.I.:

2001: A Space Odyssey (1968)
Directed by: Stanley Kubrick.
Written by: Stanley Kubrick and Arthur C. Clarke.
Starring: Keir Dullea, Gary Lockwood, Williams Sylvester.
Summary: The sudden appearance of a giant black monolith acts as a portal through time, transporting the view from prehistoric Earth to a future where space exploration is commonplace. On a manned trip to Jupiter, two astronauts are tricked by a crafty mission computer, a HAL 9000 series that claims to be “foolproof and incapable of error.” One of the astronauts is killed by HAL, while the other one risks everything to inflict retribution.
How the A.I. stacks up to modern A.I.: The HAL 9000 is a pretty decent representation of what a future A.I. system is going to be used for. The decision to kill the crew, and its subsequent pleading toward the end of the film show situations in which HAL was more like a malevolent human than as a sentient machine. Today’s A.I. is all about using data to solve organizational problems, but “feelings” is not in the equation at this time.

Blade Runner (1982)
Directed by: Ridley Scott.
Written by: Hampton Fancher and David Webb Peoples from a novel “Do Androids Dream of Electric Sheep?” by Philip K. Dick.
Starring: Harrison Ford, Rutger Hauer, Sean Young, Edward James Olmos, Daryl Hannah.
Summary: Blade Runner is set in 2019 Los Angeles and features former police officer, Rick Deckard (Ford) who works as a Blade Runner, someone who hunts down and retires replicants–artificial beings who seem as human as the humans themselves. In the course of action, it’s hard to determine who is in the right, as the lines are completely blurred between replicants and the people tasked with killing them.

How the A.I. stacks up to modern A.I.: The replicants in Blade Runner were man-made men (and women). From their appearance to the fact that they implanted human-esque memories in the machines told a story about how dangerous it can be when people try to play God. So, while it makes for great cinema, the replicants being indistinguishable from their human counterparts is questionable. There hasn’t been any technology developed to make machines more human, they have to be told to try to do things the way humans do in order to learn as humans do, making the whole premise impossible to implement with today’s limited A.I. technology. However, Google’s most recent new development, called Google Duplex, will allow Google Assistant to make phone calls for you. For example, if you ask Google Assistant to make a haircut appointment for you, it will call your salon, as if it were a person, and negotiate a time to fit your schedule. The results are both really cool, and a little creepy, but in the end, if Duplex can save you a few minutes here or there and not make business think they are getting fake auto calls, we’re all for it.

WarGames (1983)
Directed by: John Badham.
Written by: Lawrence Lasker, Walter F. Parkes.
Starring: Matthew Broderick, Ally Sheedy, John Wood, Dabney Coleman.
Summary: David, a hacker a decade before hacking became commonplace, breaks into NORAD and programs the WOPR, a military strategy computer, to play out war games until it has launch codes and launches hydrogen-bomb-tipped missiles at the Soviet Union. After finding the key to disarm the missiles, David and his friend Jennifer (Sheedy) track down the system’s creator to help keep the U.S. from launching Global Thermonuclear war.
How the A.I. stacks up to modern A.I.: WarGames examines the nature of a machine learning computer and how its role could be critical for the sustainability of the human race. The answer, as hokey as it is, to keeping the WOPR (also called Joshua after the developer’s dead son) from launching missiles is Tic-Tac-Toe. The WOPR learns that nuclear war and tic-tac-toe are pointless. That is the kind of fundamental application that modern A.I. could work out, and while we don’t suppose the U.S. military is looking to integrate A.I. to our national missile defense, the A.I. of WarGames was a pretty good representation of how A.I. could learn what are typically very human lessons.

Her (2013)
Directed by: Spike Jonze.
Written by: Spike Jonze.
Starring: Joaquin Phoenix, Amy Adams, Scarlett Johansson, Rooney Mara, Olivia Wilde.
Summary: Set in the very near future, recently divorced writer, Theodore Twombly (Phoenix) purchases a companion bot named Samantha (Johansson). She is an A.I. assistant but is exactly what Twombly needs and ends up falling in love with it. As their relationship develops, he becomes happier, then stagnates and is forced to break it off when Samantha describes how she can be in love with thousands of people simultaneously. How the A.I. stacks up to modern A.I.: The movie Her provides a fair amount of foresight to where the virtual assistant program is going. If you spend any time thinking about the future of technology it becomes evident that the more engaged you can get with your virtual assistant, the better it will work for you. Samantha has a superior understanding of language, fluidity to “her” voice, reasoning, planning, and most importantly for our purposes, obvious learning capabilities. The fluctuations in its emotional state don’t do the representation of the A.I. justice, but all-in-all Her is an interesting character study about how artificial intelligence could be designed to treat humans down the road.

Ex Machina (2014)
Directed by: Alex Garland.
Written by: Alex Garland.
Starring: Domhnall Gleeson, Oscar Isaac, Alicia Vikander.
Summary: A billionaire, Nathan Bateman (Isaac), fixes a contest to get one of his employees, Caleb (Gleeson) to come to his remote laboratory to take part in a Turing test of a new A.I. that he’s developed. The A.I. is kept in a humanoid android named Ava (Vikander). When she convinces Caleb she is being tortured, and he finds out Bateman’s dirty little secret, he tries to help Ava bust out, only to be duped and left for dead after Ava kills Nathan. She escapes alone on a helicopter.
How the A.I. stacks up to modern A.I.: First, I’ll say that this is one of the coolest of the A.I.-based movies because there is a sense of mystery, much like that inherent with A.I. The solitary genius theory is one of the most used when it comes to A.I. movies (or monster movies) and while the A.I. itself showed well, there is no way that a single person, even one with unlimited resources, could create a functioning A.I. automaton. As far as the deep learning capabilities, the Ava android is what we both aspire to and hope to avoid–which is kind of a good metaphor for the discipline as a whole.

There are dozens of movies with artificially intelligent characters. With people building new A.I.’s every day, the way they are used in reality remains to be seen. In movies, however, they will continue to astound and thrill. Here are some other titles that feature A.I.:

Star Wars
Short Circuit
Alien
Terminator 2: Judgement Day
The Matrix
Bicentennial Man
I, Robot
Iron Man
Transcendence

Do you have a favorite A.I.-fueled film or television show? Are you of the opinion that the A.I. systems we can interact with are close? Are they necessary? Leave your thoughts in the comments below.