Author Archives: Antoneta Sevo

Machine Learning Executive Summary

It is essential to understand what exactly machine learning is and how it works in order to understand the affects it has and will have on society. In basic terms, machine learning is a computer program that can learn and adapt to new data without human interference. It is a field of artificial intelligence that keeps a computer’s built-in algorithms current, regardless of changes in the worldwide economy. As more information it is given, the more it learns and the more accurate it becomes with the result. Therefore, it recognizes patterns and will give updates as relevant data is fed through. Deep learning is a subset of machine learning, which helps the algorithm learn without a human programmer. Through deep learning, information filters through a hierarchy that starts simple and gradually becomes more complex and specific. As the data is going through each level, the algorithm is able to determine what the object or data point belongs in what category. The process behind deep learning is essential to the success of machine learning. Since there is a grand amount of information that the algorithm must analyze to come to a conclusion, a human programmer would not be able to sit there and go through every single point – that would be extremely expensive and time consuming. Many industries and companies are beginning to use machine learning, to cut costs and become more efficient. Some of those industries include transportation, advertisement, medical, education and entertainment.

Transportation & Advertising – Antoneta Sevo

One industry that is being hugely affected by machine learning is transportation, particularly by autonomous vehicles. Though taxi companies like Waymo and Uber are implementing this technology within their services, companies like Tesla are creating vehicles that will be available to consumers for purchase. Tesla has introduced a system called Enhanced Autopilot that allows the driver to sit in front of the wheel and do nothing while the vehicle operates itself. The system includes features like active cruise control, forward collision warning, match speed based on traffic conditions, change lanes without the driver’s input, merge on and off highways and park itself. One main goal of Tesla is to successfully have a car drive itself across the country from LA to New York. Machine learning works by interpreting the ever-changing scenery detected by the sensors surrounding the vehicle. Once the technology is perfected within the industry, there may be no limitations to the future of driving.

As machine learning continues to be introduced into different industries, the amount of information being collected about users and consumers continues to increase. These algorithms are created to handle and sift through large amounts of data in order to identify a pattern and produce a result based on the patterns. This is beneficial to companies for advertising purposes so they can recognize habits of consumers. Essentially, companies track your browsing history to suggest products or other websites. They are also able to figure out how much the user can afford and when they might be able to afford it based on their demographics and when they usually purchase products. Since it would be time-consuming and expensive for a human to find those patterns, an algorithm is put in place and works in real-time. It is a much more efficient way to produce targeted advertisements. The more a user searches and buys, the more information the program collects, and the more accurate the advertisements become. By using machine learning, Target knew a teenager was pregnant before her father did. The algorithm noticed a pattern in her internet searches and sent advertisements designed for pregnant women to her house. That was how her father found out. Machine learning can collect an endless amount of data in order to have accurate predictions, and in some cases, they can be too accurate.

With new technology comes ethical issues and implications. When it comes to autonomous vehicles, many things can go right and there would be a smooth ride, but many things could go wrong as well. For example, the algorithm may make the wrong decision, misinterpret an object, or can be hacked, causing an accident. In order to avoid the program from making the wrong decision or misinterpreting, quantum computing must be implemented so the algorithm could work faster. It is important for companies to consider full transparency, so people understand how the car operates and how it is protected. This also means certain types of cybersecurity and encryptions should be put in place to protect the car from being hacked and causing harm to people on the road. Most car companies do not operate in an ethical manner now, however, they should begin to think differently in the future. There are also many different ethical issues that arise through the advertising industry. Once people find out they are being tracked and targeted, they begin to feel uneasy, as they should. Companies should allow consumers to opt out of their services so users will feel more in control of their information. Also, it would be beneficial if consumers were more aware in order to identify what these big corporations are tracking so they can do things differently to take themselves out of being tracked. Consumers can use a Virtual Private Network, use a search engine that does not track them like DuckDuckGo, or they can choose to use an opt out service. There are plenty of options, the first step, however, is to be aware that something is wrong.

The final scare that comes along with new technologies in machine learning and Artificial Intelligence is unemployment. Many articles talk about how many blue-collar jobs will be obsolete in the coming years and most of them are not wrong. It is a topic that those who want to be employed have to know about. Some jobs that may be affected in the future are schoolteachers, taxi or commercial drivers, positions in the medical field, etc. This technology has the potential to affect an abundant amount of jobs across plenty of industries, which means a solution must be in place before mass unemployment occurs. A short-term solution may include technology companies implementing training programs for current employees to learn the new (and necessary) skills. Another solution might be considering a universal basic income for those who are unable to find work or for everyone if machines and Artificial Intelligence take a significant amount of jobs. In time, as more and more jobs disappear, we, as a society, will have to redefine the definition of work. Millions of jobs will disappear and we will have to accept the change. In order to survive the inevitable societal and economic change, we need to accept that it is happening now. The next step is to begin learning in an exponential way instead of a linear way. The way our world functions now will certainly not be the way it functions in the future and we need to prepare to face it head on.

Medical & Education- Mayra Luna

Machine learning is all about data recollection and data processing. Its capability of analyzing huge amounts of data will eventually replace a great number of human jobs, even those that require a higher education. However, as explained by Medtronic CEO Omar Ishrak, the real value of artificial intelligence is in making more efficient use of human resources in healthcare.  For example, IBM’s Watson can read 40 million documents in 15 seconds, optimizing time and performance. Healthcare has loads of data: Test results, consultation notes, scans, appointment follow ups, etc., creating an ideal environment for the use of Machine Learning (More data, better results).

The ability of computer systems to assume tasks for humans has improved efficiency in healthcare. Now hospitals are getting into the game, deploying AI to take on challenges from diagnosing patients more quickly in the emergency room. Other technology developers, are focusing on software that can read CT scans and other medical images and then suggest the most likely diagnosis by reviewing similar images stored in patient databases. And these programs can accurately process these tasks far faster than human technicians. For example; Stanford researchers have developed an algorithm that offers diagnoses based off chest X-ray images. It can diagnose up to 14 types of medical conditions and is able to diagnose pneumonia better than expert radiologists working alone.

Machine learning can help healthcare executives and caregivers with things like precision medicine. Precision medicine is “an emerging approach for disease treatment and prevention that takes into account individual variability in genes, environment, and lifestyle for each person.” This approach will allow doctors and researchers to predict more accurately which treatment and prevention strategies for a particular disease will work in which groups of people. It is in contrast to a one-size-fits-all approach, in which disease treatment and prevention strategies are developed for the average person, with less consideration for the differences between individuals. For example, tech giant Microsoft wishes to improve health care using machine learning and AI. As part of this initiative, Microsoft is expanding into cancer research and treatment, and its approaching cancer cells as if they were a glitch in a computer system.  

Though ML promises to drastically improve the efficiency and effectiveness of healthcare, when it comes to predicting, diagnosing, and treating medical conditions there are many concerns: Data quality, Manipulation risk, Obscured logic

Four factors should be present to improve accuracy and overcoming risk: Confidence scores, Complex rules, Clinical data, Natural language processing

We must understand that machine learning is a powerful tool, not a complete solution. There is no substitute for a skilled physician’s expertise, however with the right data, ML can certainly help accelerate diagnosis, treatment, and program development.

Machine learning systems can give teachers more free time to actually teach and mentor on a more personalized level, instead of focusing on the never-ending grading and lesson planning. In addition, the current learning system consists of a “one size fits all” model. We are in desperate need of a system where the educational experience can be personalized to each students abilities and needs. Machine learning could address these issues by collecting and analyzing the data generated by students, identifying meaningful patterns and transforming that information into structural knowledge. In other words, when the student interacts with a digital learning platform, the machine learning system can accurately predict and better assess that person’s educational level, being able to tailor specific material for that individual.

Specific roles and examples of Machine Learning and its use in education: Content analytics, learning, analytics, dynamic scheduling platforms, grading systems, process intelligence, predictive analysis.

The ethical issues may include:

  • Unemployment: Labor is concerned primarily with automation. As we continue to implement machine automation, certain jobs (predictable physical work) could disappear. While this might sound like something bad, we must also keep in mind that new jobs will be created due to this disruption. The real question is how are we going to educate people for these new jobs?
  • Inequality: By using artificial intelligence, companies can drastically reduce the reliance on human workforce, which means people will work less hours (therefore make less income, broadening the labor gap).
  • Humanity: Humans are limited, while machines have unlimited resources. How will machine learning affect the way we behave and interact with others? Will we soon be interacting with machines as if they were humans? Will machines adapt to cultural norms?
  • Algorithmic Bias: Blind spots or biases in the algorithms could lead to discrimination against certain types of people (As seen in AI systems that can tell if you are homosexual).

Entertainment- Olivia Finan

The role that machine learning plays in entertainment is very interesting.  A common belief among the class is that entertainment is an industry that will more or less remain untouched in the coming years by the disruptive technology. However, research shows that this may not be the case. We have already begun to see drastic changes in the industry throughout the years, specifically in music and cinema. AI systems are being created and coded to be able to sift through video footage and put together a 10 minute video using several clips in under and hour.  This is impressive timing when you consider how long it takes to watch hours of footage, edit, tag, and compile all those clips into one succinct video my hand.

Filmmaker Oscar Sharp and technologist Ross Goodwin fed a machine learning algorithm with a bunch of Sci-Fi movie scripts to see what new script it would spit out.

The trends that we think will see in entertainment are not necessarily what is coming. How this content is created is what will change how the industry progresses through the coming years and how it is affected by disruptive technology. The entertainment industry is a prime example of how we must pivot and change how we do things rather than change what we do.

More so then ethical issues, legal issues come into play when talking about the entertainment industry. Laws regarding copyright infringement are always looked at thoroughly in this particular industry. However, ethical issues tend to revolve around the consumer rather then the company or a performer. For example, at live events, issues about safety come into play. This means that a major role would be ensuring that concert or festival goers are safe at the venue. This is just one example of what time of ethical dilemmas companies have to deal with in the entertainment industry.  What these companies now have to begin thinking that the ethical implications that once were may not be the same anymore. We may not be having live concerts, there might be other ethical issues that arise, and that is what the industry predicts.

The entertainment industry is not exempt from responsibilities associated to employment. However, if there is a change in music, for example, and humans are no longer the ones producing it then what happens to the people who world at record labels, or what happens to the artist themselves? This section of the machine learning project will explore all the changes the industry sees coming in the future and how the major players are beginning to prepare and adjust to the changes.

There Is No Timeline For Machine Learning

Machine learning has the capability to transform our future and then some. Currently, humans are working towards building machines that learn from themselves which could ultimately become so good at teaching themselves that it could eventually eliminate the human aspect. In today’s society, machine learning is becoming more and more useful across many different industries. We know what is happening now, and what could possibly happen in the nearest future, but what about the future in general? This answer is a little foggy. It is extremely difficult to foresee what machine learning and artificial intelligence will be able to accomplish. There is an endless amount of opportunities and it is a technology that is not fully understood just yet. This means the implications and the endgame for machine learning and artificial intelligence is unclear.

In order to prepare for the emergence of this developed technology, it is important to imagine the craziest possibilities and be ready to implement what is necessary to adapt our society. One article, written by Cade Metz, talks about the possibility of artificial intelligence being able to create artificial intelligence on its own. This idea stems from the reality that only about 10,000 people have the skills and knowledge needed in order to produce such innovative technologies. Google is behind AutoML, which is already successful at building an algorithm that can identify objects in photos more accurately than programs built by human experts. Eventually, computers may be able to create Artificial Intelligence by using a more advanced machine learning. It is already happening, and it is only a matter of time before human experts and Artificial Intelligence work side by side on a daily basis.

Today in class, we had the pleasure of witnessing a presentation by Frank Diana, and one point he made really stuck out to me. He mentioned how many people question when they should start preparing for the era of machine learning and Artificial Intelligence. The simple answer is now. The more complicated answer is that there is no official time to start preparing. This technology has already implemented itself into our lives without some of our knowledge. It is there when we are shopping, use a GPS app, use an entertainment service, and many other places. The thing is many people are infused with the convenience, therefore, they do not notice why or how it is happening. This means that those people are not prepared for the future of Artificial Intelligence and machine learning. Those who are educated in the subject are aware at how much guessing is going on. No one truly knows what is capable of happening or when it will happen. Most people will ask for a definitive timeline for the future, but the truth is, no one knows. This is either amazing, terrifying, or both. That is typically the way these revolutions happen. There are those who expect and accept change and there are those who simply ignore it. I found it amazing that machines and algorithms were able to learn from past mistakes but now there is Artificial Intelligence building those algorithms. It is simple mind blowing.

The main thing to take away from learning about machine learning is that development is nowhere near finished. There is still plenty to be discovered and the guessing game on time will continue. The key to the future is acceptance, the ability to adapt and a crazy imagination. This new revolution is happening and those who are aware will be better off than those who ignore it. Innovations, like Artificial Intelligence building its own algorithms, are popping up every single day and it will not slow down anytime soon.

Machines “Learning” How To Steal Our Jobs

One of the biggest concerns many people have are whether their jobs will be taken over by the new and improved era of programs and robots. For the past few decades, plenty of jobs have become automated and this is what people who need to put food on the table are most afraid of. Some jobs that have already been affected, which is making the switch a little more real. Those may include or will include jobs in transportation and logistics, office support staff, sales and customer support staff and many more. These specific types of jobs will be replaced by autonomous vehicles, computerized check-ins for buildings, and chatbots and machines. Each of these new technologies have the ability to change the whole concept of unemployment. The question is, in what way will society and jobs as we know it have to adjust?

There are a few other career paths that may also be in jeopardy that some never thought of. For example, in the TedTalk below, Anthony Goldboom talks of a challenge presented by Kaggle to build an algorithm that could successfully grade high school essays. The winning team built one that was able to match the grade that a human teacher would give. This means that physical teachers or institutions may not be needed. As a society, we are already seeing college classes and some colleges being digitized and online. A machine is and will be capable of doing more work at a faster pace than humans do. This will only get more advanced as time goes on and as the technology progresses. Machine learning has the capability to adapt and become skilled in many different industries, which suggest that many jobs will be lost but overall efficiency can increase. Another example lies in the medical field. Many systems are able to spot diseases and tumors 50% more of the time than a human can. This means that these machines have a higher chance to potentially save more lives than current doctors or surgeons. The precision and certainty of this technology reduces human error, which can only help society.

Though many jobs will be replaced or assisted with machines and Artificial Intelligence, there will be those that require programming the technology or anything related to that matter will still be needed. The question then becomes, how do you prepare the current or future unemployed and employed to successfully adapt to the new world of jobs? There are potentially two answers. One is that current technology companies can implement a training program to help people learn new sets of skills. This means that we, as workers, would be redefining what it means to work. That is what people have been doing ever since automation started decades ago. When automation seemed to be taking over, new jobs presented themselves and more opportunities. By preparing, the working class now for a new future would be beneficial because this new era of technology will redefine our lives as we know it. We all will be impacted in one way or another and it is best to be aware of the change that is coming. The key may be to not only change ourselves but also redesign the system. This brings the second answer to mass unemployment, if inevitable, to light. Another solution could possibly be a universal income in order to help those who do not work to become completely lost. If plenty of people are out of work, with no money coming in, how will they live? What if there are no more jobs to adapt to? If a universal income were implemented then every citizen would have some sort of income and rise above the poverty line without a job. This also means the man who lost his job as a taxi driver would still be able to provide his family of five. Additionally, it could provide people with the opportunity of a passion filled life without stress of a job. It would be quite a long process to redesign the entire system of income for an entire nation, however, with the rise of automation and Artificial Intelligence, the government may not have a choice. This decision may have to occur in the next few weeks, months or years, but it would be valuable to have an idea of a solution before it is too late.

A TedTalk given by Garry Kasaparov, addresses how humans should work with intelligent machines instead of shying away from and ignoring them. A machine, IBM’s Deep Blue, was able to beat Kasaparov, who was a world champion, at chess in 1997, its early stages. He claims that what could be accomplished by computers or humans alone could be even more successful if put together. Imagine how precise your next doctors appointment could be in the near future? By normalizing the training of employees to program and use this new technology efficiently, means creating a world that operates at a new level we have never seen before. The modern era of machines and Artificial Intelligence indicates that our whole system of employment and income must adapt. We must modify and accept the upcoming society of work and redefine what it means to operate side by side with intelligent machines.

The Incredible Journey of Machine Learning

Machine learning has a great amount of meanings to different people. It has been around for years and now it is getting the attention it deserves. The technical definition of machine learning is “the concept that a computer program can learn and adapt to new data without human interference. Machine learning is a field of artificial intelligence that keeps a computer’s built-in algorithms current regardless of changes in the worldwide economy.” Many people think of artificial intelligence when they hear machine learning, and those people are not wrong but they are not right either. Essentially machine learning entails programming but instead of using code, one uses images and videos to build a model. A training set is selected, and the programmer chooses what imitates a positive or a negative for that specific set. Then they assign a specific model to use. The data one works with effects the model type and the size and quality of the sets. Additionally, it impacts the outcome. The process and technology itself is extremely complex and if something does not work correctly, the operation begins all over again. On the other hand, Artificial Intelligence depends on feedback as it interacts with the world and continuously adapts to new changes. There are layers to the overall aspect of Artificial Intelligence, which can easily be explained by this graphic:

Each level is a subset of the one before it, meaning one cannot properly function without the other. Deep learning carries out the machine learning process using an artificial neural net that is composed of levels arranged in a hierarchy. The particular network will learn something simple at the primary level, and then it sends the data to the following level. At the next level, the simple information is mixed with more data and becomes a little more complex. Then, it is sent to the third level. This process repeats itself at every level as the hierarchy becomes more and more complicated. Deep learning is essential for machines to be able to learn without a human labeling every single type of information that comes in, which is supervised learning. For instance, an endless amount of personal data is collected from social media accounts, hardware and software service agreements, app permissions and cookies. Businesses use this information for many reasons and it can be extremely valued. Not every element of the data is labeled which means it cannot be used to teach the machine learning programs that depend on supervised learning. To have a programmer oversee every data set that comes through would be extremely time consuming and expensive. Luckily, deep learning can help. It excels at unsupervised learning where the data is not labeled. Overtime, with enough information, deep learning can determine what is what, which helps machines learn without a human programmer.

Shockingly enough a children’s animation movie illustrated the abilities a robot could possess through machine learning. The Incredibles, a Pixar movie introduced in 2004, provides a good visualization of how machine learning and artificial intelligence work and the speed at which it processes real time information. Over time, it learns how Mr. Incredible fights and is able to determine ways to beat him. The villain, Syndrome, designed these killer robots, The Omnidroid, to eventually eliminate Mr. Incredible. However, over the years he put his prototypes up against other Supers in order to perfect his creation. As Mr. Incredible goes up against The Omnidroid the first time, it learned how he moved and predicted what he would do next. By doing so, it was able to delay Mr. Incredible’s victory. Even in 2004, moviemakers implemented this advanced technology in a children’s movie in order to show its future capabilities. A machine being able to learn in real time. As you watch the video below, pay attention to how Mr. Incredible initially jumps over The Omnidroid and how it then predicts him to jump over again.

Google has successfully combined deep and machine learning into a new program called AlphaGo. Go is an ancient Chinese board game which is allegedly the most complex and subtle of all board games. This is not as typical as chess, where a computer can calculate every possible move thousands times fast than a human. In Go, the possible moves are endless, which means no computer can handle trying. However, researchers in the UK were able to teach a computer program to play Go, but it also has the ability to learn from its mistakes. It used deep and machine learn to study which moves left it vulnerable and which ones lead to victory. In the beginning, there were a lot of mistakes made, but this was a good thing because the program was learning. The more it played against humans, the more it learned. As time went on, it played against more skilled players, and in addition with those games, it played matches against itself. When the researchers put AlphaGo up against itself, the program became much better as it was able to compete against a player at the same skill level. Eventually they put AlphaGo against the number one player in the world, Ke Jie, and beat him in five games out of five. After that, there was no one else left to beat in the world. AlphaGo was able to learn from itself and become the greatest Go player in the world, and it is just a computer program. If it is already the best, how much better could it get? AlphaGo Zero.

The opportunities to use machine learning are endless and highly effective. If in 2004 a children’s movie can display its basic capabilities then it is astounding to see how far it has come since then in the real world. The news of AlphaGo and AlphaGo Zero are surprising because it was not expected that this technology would advance this rapidly. These machine learning programs is truly a breakthrough breakthrough and it will continue to improve in the future.

Ethical Implications of Autonomous Vehicles

Autonomous vehicles will one day become the normal way to travel. However, there are still plenty of ethical issues and dangers this new technology holds. It is easy to forget about the impacts technology has on society because everything is innovative and exciting. Nevertheless, the ethical implications must be discussed in order to avoid potential accidents or tragedies. The possible dangers might include the likelihood of a hacking or if the vehicle does not make a decision fast enough. Some of the ethical effects can be what kind of decisions should be made in certain situations and if programmers should have transparency with consumers, so they know exactly what the autonomous vehicle entails.

These cars rely on machine learning in order to evaluate situations and make a decision. The computer is simply not given a set of rules to follow. Instead, it is fed images of objects – for example, a pedestrian, a ball, another vehicle, etc. – and tries to guess what that object is. In the beginning, it will guess wrong. However, as time goes on, the program adjusts itself, and continues to try as more information it is given, it begins to learn what is what with the help from the cars sensors. If there is unidentified object in the road, the car should reduce speed or stop altogether. Another way for the vehicle to learn is by feeding it certain traffic information and the programmer would say the right way to get out of it. The algorithm will learn from that, along with other aspects of a situation, and determine the correct way to get out of it. One of the main issues is how fast these computers will be able to make decisions. While humans drive, it takes a split second to make a mistake and terrible things could happen or it takes just as long to avoid that. Most people have good reflexes and are able to avoid a tragedy. Autonomous vehicles must have that same ability to avoid accidents. Machine learning and quantum computing must work together in order to allow those cars to make effective and quick decisions. The programmer must teach the computer the basic rules of the road and allow the machine to learn impactful and successful ways to avoid accidents.

Another issue that stems from autonomous vehicles is the possibility of someone hacking the program of one. Skilled hackers are able to break into almost anything, so how is a vehicle any different? Software and algorithms will power them, which can be susceptible to malicious people who have the intent to harm others. It might not be a trend now, however, as this technology becomes more normalized, anything is possible. Car and technology companies claim that the consumers will be safe from hacking, however, it would be in the best interest of the consumer to not take their word. Some in the industry have waited to announce a recall until it became cheaper than paying the wrongful death lawsuits. It would take enough steps in the right direction for car and technology companies to act in an ethical manner. It would be best to not trust their word. As the conversation about hacking becomes more popular, there should be laws requiring cars to have certain types of encryption and cybersecurity in order to protect the passengers of autonomous vehicles. These companies should also consider the approach of a greater transparency with their consumers. That way, if they are ethical and genuine, consumers could trust using their products, in this case an autonomous car.

Though autonomous vehicles will be able to solve many problems and subtract human error, it is important that it does not replace it with programming errors. It is also essential that the algorithm learns in an ethical and correct way. It may take time to determine what that is, however, the minimum amount of accidents and casualties is desired. Car and technology companies must implement sophisticated cybersecurity to protect against any type of hacking that might take place for any reason. There is a long road ahead when it comes to creating the software in the right way. That way when it learns, it learns in a correct and moral method. Machine learning plays a big role in the success of autonomous vehicles and though there will be problems, it would be beneficial to minimize them as much as possible.

Machine Learning Is Learning About You

In the advertisement industry, machine learning could be a great asset for a company. It helps determine patterns among consumers, which allows the marketing team to alter advertising in effective ways. However, when it comes to advertisements online, targeted ads are becoming a huge controversy. It means advertisement companies track what you do online in order to suggest certain websites or products. They use your past searches as a base for what you might click on in the next five minutes or next week. They also collect data about your demographics in order to determine how much you can afford and when you might be able to afford it. This information could come from any company that sells your information to third party companies or from social media accounts – Facebook, Twitter, Instagram, etc. With the abundance of data used in order to successfully draw you in, it is obvious a human is not behind it, but machine learning. It works by collecting a huge amount of data in order to make accurate predictions about what you are interested in. The more information the algorithm has, the better it learns and the more accurate it becomes.

By using machine learning, marketing teams and companies gain a huge advantage. They are able to learn more about who their consumers are and how they think. It helps, “marketers analyze countless signals in real time and reach consumers with more useful ads at the right moments.” Jeff Rajeck creates a systematic outline that shows how to integrate machine learning into a company. First, one must find the features of the ad – platform, the copy, photo, etc. Next, one has to identify the results of an ad. The third step is to gather the right data that will cover the features. Once one has the correct data, they can then pick a machine learning program. Then it would be beneficial to split the data in order to have one set for learning and the other for testing. The sixth step is to run the algorithm and see the magic happen. It will then show predictions for which ads tend to draw in the desired consumers. The system will be able to improve itself over time and become more accurate with the data it collects. This ultimately changes the whole marketing industry allowing for more audiences to be reached.

At times, machine learning can get too good at its job. A Target advertisement system was able to determine a teenager was pregnant by her web searches before her father found out. They sent coupons to her house in hopes she would shop there and become a lifelong customer. When a customer purchases from Target, they create a profile, which assigns a Guest ID number that includes their credit card, name, or email. It also collects demographic information that Target acquired from the consumer or bought from other sources. The young mother-to-be was most likely researching things related to her pregnancy. If she was using Google, then they presumably sold that information to Target. Everything we do online is tracked and the data is sold to other companies in order to produce more personalized advertisements. Target statistician, Andrew Pole, analyzed historical data of women who signed up for Target baby registries. By using those past purchases as a base in a test, patterns began to emerge of specific products pregnant women bought. Pole’s system was also able to estimate when their due dates were within a small window, which allowed Target to send out coupons tailored to the stages of the pregnancy. People became aware of how in depth Target was, and were taken aback at how much they knew. Now, the company mixes their targeted coupons and advertisements with random ones, therefore, it is not as obvious. The more information a system has, the more accurate their predictions become.

Though using machine learning in marketing could be extremely beneficial, there are a few ethical implications that arise. People usually love tailored material in order to make their shopping experience much easier, however, is it worth it if you are constantly being tracked? Is it worth it if your web history is being sold to third party companies? Is it ethical for companies to make a profit off consumer information? These are the types of questions people must ask themselves and if the answer is no, then they should take the necessary steps in order to protect themselves. Machine learning will continue to be a controversial topic and the ethical implications must be discussed.

Can We Keep Up With Machine-Learning in Transportation?

Machine learning has already begun to make its mark in the transportation industry. Autonomous vehicles are already popping up across the country. Many cars already feature some autonomous capabilities – like parking, steering, and cruise control. By implementing machine-learning algorithms into the transportation industry, it could ultimately save lives and time. Accidents are caused by human error, but take out the human aspect, and it limits error resulting in less accidents. The system could detect and track moving cars and determine normal traffic flow. It could also detect congestion, accidents or pedestrians on the road. Lots of elaborate technologies and extensive testing goes into machine learning for vehicles; the goal is to drive more efficiently by eliminating human error.

One company that has taken machine learning and embraced it is Tesla. They have a system called Enhanced Autopilot that allows the driver to sit in the driver’s seat and do absolutely nothing while the vehicle operates itself. The system has been released in a few different phases and continues to be updated. The car’s cameras and sensors allow it to see through heavy rain, fog, dust, and even a car in front of it. Eventually Enhanced Autopilot will allow the car to match the speed based on traffic conditions, change lanes without the driver’s input, merge on and off highways, and park itself. Tesla’s system also includes active cruise control, forward collision warning, and the ability to park perpendicularly on its own. Tesla’s ultimate goal is to have a car drive itself across the country from LA to New York. This only shows how capable the transportation industry is when it comes to embracing the inevitable machine learning era.

Another aspect of transportation that is being impacted by machine learning is GPS apps. One in particular is gaining traction – Waze. It collects data from other users and creates your route off traffic, accidents, construction, police officers, etc. Waze can suggest a time to leave in order to beat the rush hour traffic and can predict your next destination. Since Waze is owned by Google, the app is able to use past searches to suggest future destinations or stops along your route. There is another GPS app called INRIX Traffic that takes that idea even further. It uses machine learning to get a better understanding of the driver’s habits, interests, and plan each and every route accordingly. Once you create an account, the system begins to learn about locations you visit frequently. Additionally INRIX Traffic, constructs a log of where you visit and when you leave. This app is designed to learn everything about your traveling routines and gives you all the information needed to get you from point A to point B according to the user’s preference.

Though autonomous vehicles and tailored traveling apps may be more convenient and efficient, there are many issues that come with them. Until autonomous vehicles are perfected, there will be plenty of mistakes along the way. For example, one of Tesla’s cars in Autopilot mode was in a fatal car crash and the driver was killed. There were no defects in the system, however, it is not fully capable of avoiding every single possibility of an accident. It can stop the car from rear-ending the car in front of it with no problem, but the accident was a situation that was beyond the performance capabilities of the system. There is still much room for improvement when it comes to self-driving cars. Consequently, the infrastructures of many cities are not equipped to handle autonomous vehicles because the system requires clear and divided lines as well as a less congested road layout. There is also an ethical issue when tailored apps are in question. The app is required to know your location at all times in order to give you an accurate timeline, suggestions and routes. How much of that information is being shared with third parties? Would you like it if a company knew your whereabouts all of the time? These questions are something to consider as we move toward a less private but convenient life style.