Georg Tichy

Created 0 Campaigns

stratecta

Gender Identity for AI

Artificial neural networks have given AIs the functionality for complex problem solving and pattern recognition, and they have entered the workforce, particularly in areas of big data analysis and global finance. As we begin to interact with and study these new learning machines, interesting questions arise. Are they going to take on human behavioral and gender distinctions (gender identity), because they have been programmed with data sets that have unconscious bias? Will those who are giving the learning machines feedback to focus their problem solving allow behavioral constraints into the teaching? If we give the AIs a woman’s voice, and a woman’s name, will we interact with her as if she was a woman? And does that mean she will in turn internalize those social expectations and become more female?

Naturally we are interested in all things having to do with gender. It is the first sentence the world places upon us, when the midwife announces boy or girl. We love gender. We give our teddy bears genders, and can describe in detail why we think-no, why we know that our little darling is a boy or girl. We give our cars genders, names, and personalities. It’s just because we’re human, and we want to humanize the things we love, and that surround us. And part of humanizing inanimate objects is to give them a name, a gender, and shower them with affection.

Part of our fascination with gender has led to some poor science, the popularity of which has trickled down into our collective consciousness. The idea that male brains and female brains are different in a significant way is probably not true, though the debate rages. Structure follows function, and hormones affect the developing brain. But even with minor structural and functional differences in the brains that are most probably hormonally-based, there is very little difference in boys and girl’s brains. There is a much wider variance between individuals than can be measured than between generalized groups based just on gender. We are more complicated than can be described in pop-science about hardwired aggression and nurture vs nature.

What is different between genders is communication, how we use language, and there the gender differences are significant enough to be measured. If we think of communication as the way we input data into our brains, we grow our biological neural networks with the complex range of human communication to which we’re exposed. And there are differences between male and female communication.

So with the science showing that biological neural networks- aka human brains- are more complex than can be measured, but are influenced by hormones, language, biology, and the wide range of human culture, we are left to consider if artificial neural networks will also be influenced by language and human culture. (This is assuming that the artificial neural networks that are biology and hormonally mediated are still a few years in the future.) Continue reading

stratecta

Digital Transformation: Things a Digital Document Can Do that Paper Can’t

When companies are wondering whether or not to go through the most basic stage of the digital transformation, moving from paperwork documents to digital document management, there are often a lot of fears about how implementation and unfamiliarity with the new system will slow down productivity and potentially confuse the staff. However, the same things can be said about any major change, including drastically altering the catered lunch menu. The real thing that administrators and business owners should be considering about the digital transformation is all the ways that digital document management can enhance the efficiency of your business both in predictable and unpredictable situations.

To help you understand the drastic difference between a business run on paper and a business run through digital documents and software, let’s narrow the focus down to the humble document. All the things that can be done with a digital document, but on for which the original and every copy is paper.

1) Same Document Form for Drafts and Final Copies

The first thing to realize is that no one writes their documents on typewriters anymore which means that nearly 100% of modern documents and paperwork start in digital form on a word processor. That word processor saves a digital document which is then printed out. Though many companies who work with paper still think of a paper copy as ‘the original’, in truth, the originals of all but historical documents are now digital. The paper is the real copy and every time an edit is made or a new version is drafted, the document is created in digital form, printed to paper, and then interacted with.

Why not just skip the paper stage? When you work with digital documents, there’s no need to print unless a client needs a physical copy for a specific reason like pen-and-ink signatures or they request a hard copy for their own private records. Otherwise, you can receive, develop, work with, and submit documents all in a single digital form.

2) Infinite Editing of a Single Document

When you’re working primarily with physical copies of your paperwork, edits are not just challenging in that they must be done carefully and neatly. Every old copy will need to be tossed in favor of new print-outs of the edited work. Edits on paper are permanent or, even with hand-written documents done in pencil, require wear and tear on both the eraser and the paper.

Digital documents, on the other hand, can be edited an infinite number of times, revised, corrected, and collaborated on without an eraser white-out/liquid-paper, or constant printing and re-printing because digital edits are easy and cost nothing. Along the same lines, the edited document and the original can be the same file, ensuring that everyone who has access now has access to the updated version. Continue reading

lawyers

Insurers and Lawyers are stepping into AI

The insurance business revolves around measuring and pricing risks. In short, it’s a business about caution. Perhaps that’s the reason why insurance companies have lagged other industries, such as banking, in adopting new technologies that offer more powerful analytics capabilities. But the insurance industry’s reluctance to adopt new technologies is breaking down. So-called insurtech is one of the industries drawing investor interest as startup companies test and market new software solutions, Coin Journal reported, citing a report from CB Insights and KPMG. Insurance companies are interested in using Internet-of-Things technologies to identify and mitigate risk, while also incorporating other technologies that identify fraud, improve efficiency, and cut costs.

Financial services quickly embraced new technology because its use in improving profits was readily apparent to bankers and investors. Gathering information in real-time and timing transactions based on the steady flow of information helps financial industry players make money. In the insurance field, the name of the game is saving money. New technology can help insurance companies more accurately price policies to risk and while also cutting down on fraudulent claims that are costly to the industry. Daily Fintech notes that one of the more transformative developments in insurtech is the emergence of telematics. The capability to gather and transmit near-real time information produces even more data points from which an insurer can more precisely make their risk assessments. These bountiful data sources produce a tremendous amount of data from the home for home insurance, and from the car for property and casualty insurance, Daily Fintech explains.

The collection and analysis of data is not limited to the home and the car. Wearable technology first found its place in people’s lives in fitness applications. But these technologies have matured to a point where they can be used for healthcare applications. The ability to monitor people and collect data of a person over a longer period of time yields measurable data that insurance companies can use to assessing health risks for life and health insurance, Daily Fintech says. Continue reading

predictive analytics

Better Planning with Predictive Analytics

In industries with longstandig customer relationships, you want to know how customer-churns are impacting the customer value. The better you can analyse the future churns within or outside your company, the better you can react to these changes. Typical industries for this analysis are: Telecommunications, Banking, etc. The ability to access and analyze data from many cultures and geographic regions is one way the current machine learning platforms can produce models that deal with global human behavior. In the past, when we were dependent on subject matter experts, predictions were often limited by geography and culture.
Predictive analytics will help you to make more meaningful analysis with Big Data and allows you to make forward-looking business decisions. Before starting a predictive analysis, you need to know the components of your value drivers in details. E.g. the churn rate is one important value driver and the financial contribution margin analysis includes the following categories:

Net Revenues
-Direct Costs
Contribution Margin 1
-Acquisition and Retention costs
Contribution Margin 2
-Network Costs
Contribution Margin 3

Customer segmentation and churn modeling are important underlying tasks in order to make relevant predictions in the telcommunication sector. See example:
• Churn within your company (within price plans): In order to predict the churn-likeliness of customers you need to take into consideration the Social Data and Usage Data of your customers. The usage data includes trends of the following pricing components: Voice-Call-Patterns, Data-Usage, Roaming, Wholesale, etc. New market-developments like new handset models need also to be incorporated in your analysis. Social Data includes social network data (comments, recommendations) of your customers. Both (social and usage data) need to be merged and analysed (time line, clusters, regression, etc.).
• Churn outside your company: To get a better accuracy of churns outside your company you need to include the pricing developments of your competitors in your analysis.

Trying to measure the success of a group of company based on a limited number of KPI will not lead to a fair, motivational decision. It will lead to an oversimplified judgement of a situation and to shortsighted conculsions.
What can be done in order to improve a decision making process of a group:
• Use as many data sources as you have (big data, data lake)
• Use external data and research data – to bring the external view in the decision making process
• Find patterns relevant for your group or company – the patterns are different for all companies (you can’t guide companies with the same five KPI).
• Look at trends and personal feedback – build a story around the data
• Use IT extensively to measure and analyse data.
• Do not measure the KPI in a traditional financial form – the new KPI is explained in multidimensional ways.
• Automated processes are welcome, but be aware, that for an holistic picture you need good human analysts.
With today’s tools you can evaluate complexity much better than years ago. It’s possible to make better decisions in a more sophisticated world. Predictive Analysis helps you bringing your Business Intelligence to the next (analysis-)level and predicts how your customers are going to react in your price plans in the near future.

Continue reading
stratecta - ransomware security

Overcoming The False Hope of a Ransomware Attack (Part 1)

Every modern business deals with a certain amount of technology. From tech companies that consist internally of nothing but professionals at computers to minimally technical industries that still rely on databases and business software to keep everything running smoothly, the need for a secure network, email security and backups of archived business data is universal. When your data is in danger and it looks like there’s a chance of recovering anything that has been lost, most companies will jump through flaming hoops for any either planned or, worse, unplanned recovery method. That is exactly why ransomware is so terrible. The hope of getting your files back after a disaster is often more powerful than the fear of losing them in the first place.

While you may think that your files are being held hostage, your disaster recovery plan is much more reliable than any hacker’s “promise” that you’ll see your files again.

Malware Has Always Wiped Files

To understand the innovation of ransomware, it may help to have a better grasp on the history of malware as a whole. Ransomware is just one of the most recent innovations in a long chain of malicious, invasive software. In fact, while there has been a significant rise in the ability of malware to actually do something like steal credit card numbers or extortion, malware has traditionally been almost completely pointlessly evil. Worms have roamed the web since before the internet unification seeking out vulnerable systems and often infected websites are simply left up to hurt anyone who comes across them.

When an infection is successful, whether it was targeted or random, the malware’s goal is simply to cause pain. Spamware makes your system unusable with constant pop-ups, spyware steals your login information and uses it for fraud or more spam, and many forms of malware despite the name will simply explore your files, deleting or corrupting them as it goes. Hackers have always deleted files for fun and there’s no reason to assume that they’re going to stop now just because they’ve also figured out how to make a little side cash.

What Ransomware Does

When ransomware gets onto your computer, it’s first act is usually to lurk around for a while. During this time, it may finish installing itself, spread from the first computer into the local network, and map all your files. These processes usually happen quietly using background resources and the delay often masks the true infection point, whether it as a bad website, a phishing email, or an actual hacker security breach in which the ransomware was placed on your computer. Continue reading

healthcare

Behavioral variables and nanobots in healthcare

Big data and machine learning platforms are in a unique position to analyze one of the most challenging aspects of medical research: behavioral variables that are not reported accurately by common assessment methods. While EMRs (electronic medical records) prompt healthcare providers to collect a great deal of subjective information that impacts healthcare, such as compliance with medication regimes or alcohol intake, the validity of the information collected is questionable.

The subjective nature of the reasons people conceal or alter information given to a health care provider are as complicated at the whole of the human population. People feel social pressures to conform and please a questioner. They don’t want to admit to money problems that impact health care. They do not accurately see their own behavior. Cultural norms regarding personal information vary widely, as does disclosure by age and gender and social class. But new methods of gathering and quantifying data across populations has the potential to give greater insights into human behavior that can change the results of medical research.

Relying solely on patient reports of behavior is a method of gathering data that is extremely limited and may significantly impact the results of healthcare research. But gathering self reports, along with subjective research reports, pharmacy records, laboratory test results, social media, buying behavior, financial records, employment records, and other sources of data, and then analyzing across populations, can give a more accurate picture of what people are actually doing. By having a more accurate picture of human behavioral variables, healthcare research can more accurately assess the impact of human behavior on health care outcomes, and propose treatment modalities that are fine-tuned to the people we actually are. Continue reading