Richard Proszowski

Richard Proszowski

Supervise and Manage Output – Not the Employee

At what point in the hierarchy of an entire organization does management shift from oversight of purpose and ensuring outcomes to management of the employee? At what point does effectively providing guidance and setting appropriate goals for employees translate to "I have to supervise you" versus "I want to ensure you are adding value to our organization in the intended manner"?

Is Computational Humor Funny?

The application of artificial intelligence and analysis to understanding and generating humorous content, such as jokes, puns, and humanly amusing stories is referred to computational humor. These area of AI development can also be called computational comedy. Of course, the first thing that comes to mind is the relative oxymoronic ring to computational comedy. Isn't comedy by its nature instinctive and more closely aligned with reaction than with computation? Maybe I just don't get it!

Deep Learning Super Sampling and Collaboration

Nvidia's new processors have Deep Learning Super Sampling (DLSS) embedded as part of the core architecture. In marketing information from Nvidia, the architecture is considered the performance multiplier for optical frame generation powered by AI. Fundamentally, the DLSS capability can "fill in the blanks" in content to create more realistic and more detailed renderings of images in real-time.

Human Memory Consolidation compared to AI Machine Learning

In terms of human memory, acquisition is the process of experiencing and storing new knowledge in memory. Memory consolidation refers to the process that occurs when a memory becomes permanent.  Lastly, recall is the process of retrieving previously stored memories. According to researchers, while acquisition and recall occur while awake, memory consolidation occurs in people when they are asleep.

Media Echo Chambers and AI

When personalization of news and information reaches a level where people are primarily exposed to information that reinforces their own underlying perspectives, a media echo chamber has been established. Fundamentally, individuals get repeatedly exposed to information that reinforces their own opinions versus presenting a diverse universe of information. If you get to pick what news you want to hear based on being aligned with the news and feeling good about the news, opinions get reinforced versus being balanced.

Show Your Work

It is significant to consider that AI models are trained on large amounts of data from the internet which may contain biases and reflects the biases of society at large. The model builder or AI trainer has the ultimate oversight in the inherent logic and perhaps perspectives that the AI inherits. I wonder if once something is "learned" by an AI model, how difficult it may be to "unlearn" if it is indeed a bias or trait not originally intended to be developed.

How will Moore’s Law Translate to AI?

Very simply, AI is the next evolution on the scale that Gordon Moore postulated in 1965. Although, at the time, Moore's law was very hardware based and transistor density centric, if one extrapolates to generic "computing capabilities", Moore's law is alive and kicking and accelerating us into the AI universe.