1. For the technological topic of your choice, say what it is, give a present-day impact on individuals, and your opinion about whether it is good, bad, or whatever you think.
Machine Learning (ML) is my topic of choice. ML is having an impact on individuals at every stratum of society. Personal digital assistants like Siri, Alexa, Google Assistant, Cortana are enabling spoken interaction recognized by ML. The ensuing conversations are then facilitated by ML and queries and predictions are processed further by ML. This untethers computing from the mouse and keyboard, further facilitating mobility. My opinion is that this is good.
2. For the technological topic of 1 above, give five more present-day impacts on individuals. One of the five should relate to your career path. For each, give a good reason why the impact exists.
i. Change of Computational Paradigm: A disruptive change that breaks free from the imperative and applicative styles of algorithm execution, facilitating a more natural pattern recognition style of computation for everyone. Not since the advent of the personal computer, or the internet has a more significant change occurred. ML is a personal technology affecting individuals.
ii. Medicine: Skin Cancer is already more accurately diagnosed by machines than by experts. This will continue into other specialties of medicine, especially those that can be reduced to image or numerically based ones, like radiology, histology, pathology and clinical laboratory analysis. This will reduce medical costs and facilitate preventative medicine. Not getting cancer affects individuals.
iii. Shopping: ML has been affecting shopping for some time for online services like Amazon, Walmart and Target. Annotations like, "People who bought x, also bought y, and z" are produced by ML. Recommendations are similarly generated. These services affect individuals.
iv. Transportation: ML is currently part of a space race between companies like Uber, Lyft, Waymo and Google for creating driverless vehicles, and for the more mundane tasks like scheduling rides, matching riders to available vehicles and enabling real-time navigation and changes in the process. Getting where you are going is a personal, individual experience.
v. Entertainment: ML facilitates matching future programming to that which individuals have already consumed. This takes place on services like Netflix, the Apple Store, Spotify, Pandora and Amazon Prime. It is individuals that are entertained.
3. For each of the 6 impacts of questions 1 and 2, extend to past impacts and one possible future impact. Conciseness is acceptable...no need to write a book about this.
from Q1:
It is hard to comment on past impacts since ML is relatively new, having only existed in its present form since 2016. Because of this all my answers below with respect to past impacts will be somewhat attenuated. We can speak in terms of the technology not existing and the impact that NOT having it would have, or we can speak in terms of the impact that the technology has had, which we have already done. To address this position further we have to broaden our view to include those advances that have enabled ML to exist in its current form. I will list these advances in terms of changes of paradigms that have occurred in the hardware, software, languages, operating systems and human interface arenas.
from 2i:
For brevity I will give a few key examples from each area.
In computer science there have been several major computing changes of hardware paradigm:
- mechanical relays
- vacuum tubes
- transistors
- integrated circuits,
- VLSI
- microprocessors
- GPUs - first for graphics, later for ML
- TPUs
- machine language
- assembly language
- lexical analysis
- LL(*), LALR(1) parsers - used for Python
- recursive descent parsers
There have been major changes in computer languages:
- Fortran
- Lisp
- Algol
- Pascal
- C, C++
- Java/Javascript
- Python - the principle language of ML
There have been major changes in operating systems:
- OS360
- TOPS10
- VMS
- Unix - a common ML platform
- Windows - a common ML platform
- MacOS - a common ML platform
There have been major changes in user interface:
- Punched Cards and Printed Output
- The Paper TTY
- Calcomp Paper Plotters
- The Glass TTY
- The Mouse
- Frame Buffers
- Computer Graphics Technology and Scientific Visualization
- Bit-mapped/memory-mapped Display
- The Integrated Personal Computer
- The Smartphone - an ML Platform
- The Tablet - an ML Platform
- The Personal Digital Assistant - an ML Platform
It is worth noting that Unix has been in place since the late seventies. Machine Learning sits atop the current apex of hardware, software, languages, operating systems and user interface improvements. Computing has become inclusive of its own past and past impacts. Text processing and analysis tools developed in the late seventies remain viable today, albeit in improved form, running on faster hardware. ML will be the same way. A future impact will be the way ML changes the activity of programming itself, possibly to a conversational one.
from 2ii:
Past impacts of ML to Medicine? We are just getting started.
A future impact is illustrated by the query: "Hey Google, build a database of NIH tuberculosis radiology and run a supervised learning session to answer which lobe of the lung is affected most in children between nine and eleven years of age."
from 2iii:
Past impacts of ML to Shopping? Again, we are just getting started. But the lines between past and future are starting to blur because older services will be remodeled and repurposed using the conversational interaction that ML provides. A future impact is illustrated by the interaction:
User: "Hey Alexa, fix dinner and deliver it."
Alexa: "What do you want me to fix?"
User: "The usual."
Alexa: "Okay, your pepperoni pizza will be delivered in 30 minutes."
from 2iv:
Past impacts of ML to Transportation? There have been a couple of significant accidents with driverless cars, causing vendors to be more cautions about releasing them to the public. Ralph Nader, in response to the Boeing Max 800 accidents that killed his grandniece, insists that we will have to keep humans in the loop rather than deifying machine learning and artificially intelligent machines.
A future impact is illustrated by the interaction:
User: "Hey Google, send a taxi para-drone and a passenger harness"
Alexa: "How heavy is your passenger?"
User: "100 kilos."
Alexa: "Okay, your drone will be waiting outside in five minutes."
from 2v:
Past impacts of ML to Entertainment? Too soon to tell. A future impact will be that ML and Augmented Reality will enable:
- The invisible wall between audience member and actor to blur.
- The line between movie and theatre will blur.
- The line between computer game and science fiction program will blur.
4. For each of the 6 impacts, give a second alternative possible future impact on individuals.
We have a second possible future impact of
- ML on itself
- ML on Computation
- ML on Medicine
- ML on Shopping
- ML on Transportation
- ML on Entertainment.
from Q1: ML on itself
Unlike conventional programming, Machine Learning discovers the algorithm by being trained with training data. This algorithm discovery process will increase in sophistication and eventually ML will figure out how to discover ML algorithms.
from 2i: ML on Computation
With the need for fewer programmers, ML could result lower employment rates for those with a computer science or programming background.
from 2ii: ML on Medicine
With routine diagnosis and treatment partially automated there will be a need for fewer doctors. One doctor will be able to do the work of many, and will only be needed in ambiguous or difficult cases. As in the computer science case ML could result lower employment rates for those with a medical degree. Just as the internet eliminated the middleman, Machine learning will eliminate the domain expert.
from 2iii:
With no need for brick and mortar stores, there will be less need for cashiers, sales clerks, stockroom personnel, and managers. Physical stores will be replaced by Virtual Stores like Amazon, WalMart, and Target. These in turn will utilize ML and robotic automation further reducing the need for human staff. The main job will be repairing the robots when they break, which will be done by replacing them. These robots will be manufactured on automated assembly lines, so only the most meta-level engineers and skilled repair staff will still have jobs. Even so, producers of products that are distributed via the Virtual stores will still be needed. They will include jewelry and sand sculptures made by people who live on the beach, wear shorts, play video games and smoke pot.
from 2iv:
With driverless cars, buses, trains and aircraft, there will be no need for drivers, conductors or pilots. When these transportation systems fail, airbags will inflate and parachutes will deploy. Dazed passengers that survive the crashes will wander aimlessly, making smartphone calls and trying to find another way home.
from 2v:
With fewer jobs, the primary pastime will be entertainment. This will include binge-watching various television series, movies, and AR and VR theatrical productions and immersive video game experiences articulated above. With less reason to be concerned about quality control, many will spend most of their free hours using entertainment and recreational drugs.
5. For the technological topic of your choice, give a present-day impact on an organization, such as business, government or others, and your opinion.
Machine Learning has enabled Amazon to become the largest shopping service in the world and its owner, Jeff Bezos, to become the world's richest man.
6. Give 5 more present-day impacts on organizations. Of these 5 and the 1 you just discussed above, at least one impact should relate to business, one to government, and one to some other type of organization. For each of the 5, give at least one good reason.
i. The TensorFlow software for Machine Learning was open-sourced by Google and has made TensorFlow the defacto standard for ML development.
ii. ML is used by Facebook to perform automatic facial recognition for tagging images.
iii. ML is used by Netflix to optimize its revenue streams.
iv. ML is being used by the United States to create smart weapons.
v. ML is being used to predict stock market behavior by sentiment analysis.
7. For each of the 6 impacts of questions 5 and 6, extend to past impacts and one possible future impact.
from Q5:
As in the case for the individual, the impact of ML on businesses and governments is still in its infancy and it is too soon to tell. But there is a new battlefield opening up in cyberspace between competing nation states, vying for superiority in intelligence gathering, spying, intellectual property theft, and cyber attacks. Future wars will be fought in cyberspace, and spill over into the real world as a consequence.
from 6i:
The past impact of Google was in search, document retrieval, book and paper digitization. The future impact of Google will be in using these data sources as the fuel for ML training, testing, deployment and prediction in smart services formats as we are seeing with digital assistants.
from 6ii:
The past impact of Facebook was in bringing people together and letting them share experiences regardless of their geographic location. But they have used their collection of personal information to generate revenue and throw elections without concern for their constituency, which could lead to their future demise.
from 6iii:
The past impact of Netflix was bringing movies and television shows to the desktop, to the family room, eliminating the need to go to dangerous movie theaters, where the floors are stick, and shootings happen. The future of Netflix will continue this trend, making movie theaters a thing of the past, especially in an atmosphere of growing violence.
from 6iv:
The advent of smart weapons in the nuclear age will make wars, when the do occur, swift and final. Second and third world countries which do not possess or control significant computing resources will be unable to wage war in any but the most Luddite of fashion. The demise of ISIS is an example of the old style warfare meeting the age of the drone and intelligent system. In this case civil disobedience and domestic terrorism may become more commonplace with damage to computing infrastructure taking place as nation states and sovereign individuals engage in cyber warfare.
from 6v:
Stock market trading patterns will shift as algorithmic trading powered by machine learning causes fortunes to be exchanged on millisecond time frames.
8. For each of the 6 impacts, give a second alternative possible future impact on organizations.
from Q5:
With the reduction in employment, governments, organizations and businesses will be obligated to provide minimum basic incomes for those displaced from the workforce by Machine Learning. Educational institutions will have to convert to faster training formats to service the multiple careers that people will have as ML advances, obsoleting their previous career at each turn.
from 6i:
ML will reach beyond the informational and extend into home automation, robotics, smart appliances and services like water faucets that vend out water at a specific temperature and log when users wash their hands providing the alibi's and warnings of medical conditions like peripheral neuropathy that occurs in diabetes, detectable when a user continues to request increasing water temperatures because they can no longer feel the water on their hands.
from 6ii:
ML and blockchain will be combined to enable automatic real time voting on systems that protect users privacy. Services like Facebook at that exploited their user base will give way to more subtle exploitations and advertising that is more tailored to the user and more respectful of their privacy.
from 6iii:
Another impact of ML, in an era where personal entertainment is growing is personal exercise and workout regimes enabled by machine learning. Examples of these are virtual bicycle tours of the world where the user never has to leave their home to obtain a good workout.
from 6iv:
As governments engage in smart arms race, individuals will use computer viruses, and devices whose construction is facilitated by the abundance of information an anonymity of the blockchain to wage war against the state at a using machine learning. This will be a high tech evolution of old style guerilla war tactics.
from 6v:
Financial singularities resulting from stock market trading under the control of machine learning programs will cause unpredictable rises and fallings in the market. Naked short selling will occur when some companies are targeted by ML programs for automated takeover and gutting of their capital resources causing them to go out of business and cease operations in a single moment.
9. For the technological topic of your choice, give a present-day impact on society, and your opinion.
The impact of machine learning on transportation will eliminate the idea that individuals will own vehicles that must be parked 95% of the time.
10. Give 5 more present-day impacts on society. For each of the 5, give at least one good reason.
i.
Fewer vehicles in circulation will mean car dealerships will be things of the past.
ii.
Fewer vehicles in circulation will mean a need for fewer parking lots and fewer parking spaces.
iii.
Unused parking lots can be repurposed for new buildings, housing and recreational purposes.
iv.
Machine learning will also result in fewer stores and the value of commercial real estate will plummet as fast tracts of existing buildings are repurposed.
v.
Instead of vehicles being customized to the image of their owner, they will become fewer in kind, fewer in color and less distinguishable by decoration.
11. For each of the 6 impacts of questions 9 and 10, extend to past impacts and one possible future impact on society.
from Q9:
The last time society went through an impact of magnitude comparable to Machine Learning is when the automobile was originally invented. Streets and highways had to exist before cars and delivery trucks could go into widespread circulation. Now the infrastructure has been built, but the nature of transportation is changing. The nature of the delivery of goods and services is changing. The 'Amazon Effect' of dumpsters filled with one-use cardboard boxes will give way to nearly 100% recycling and standardization of shipping containers.
from 10i:
Even though car dealerships will eliminate their showrooms, their repair shops will remain active. But the variety of cars and the number of cars they service will drop precipitously. There will be fewer cars and trucks, but those cars and trucks that remain will have many more miles on them.
from 10ii:
Fewer parking spaces will result in cities having less revenue from parking meters. There will be no need for meter people to give tickets and those people will not have jobs. There will be fewer cars towed away and impounded because individuals will not own them. Tow trucks will still exist for those vehicles that are driven non-stop 24/7/365.
from 10iii:
With more real estate available housing costs will drop and it will be possible to afford a housing in high density population locations more easily.
from 10iv:
Fewer stores will compound the affordability of houses for people at all income levels. Income levels may bifurcate further into rich vs. poor unless guaranteed basic incomes are mandated and implemented.
from 10v:
Fleets of every kind from taxi fleets like Uber and Lyft, to delivery truck fleets, to aircraft fleets will consolidate. There will be fewer vehicle types as the economies of scale impinge on the holders of large fleets who will unify and simplify their vehicle inventory to reap the economies of scale.
12. For each of the 6 impacts, give a second alternative possible future impact on society.
from Q9:
The impact of Machine Learning on society will be as far reaching as that which occurred when the automobile was invented. Social rites of passage, like learning to drive, owning a car, going on a date in a car will end. Going on a vacation in a personal car will give rise to new group and family destination activities facilitated by entertainment companies. Experience-based gifts will replace material-gifts, since material possessions will lose value in a highly virtualized economy. In their place will be virtual collections of objects in virtual cyberspaces. These virtual collections may be copied or protected by blockchain technologies ensuring their uniqueness.
from 10i:
With fewer domain experts in circulation the need to supply large numbers of them in the industrial, engineering, medical and scientific arts will decrease. With knowledge being instantly available via Machine Learning mediated conversations with homebots, the notion of going to school for four to eight years to obtain a professional degree will change.
from 10ii:
Just as stores will decrease in number, so will schools. More people, especially in a age of random shootings by the discontent, will elect to train their children in safe home environments. Since massive online courses taught by the best teachers will prevail, there will be need for fewer trained professors and teachers. They will share the same fate as other domain experts. More people will make jewelry and sand-sculptures. They will play music, dance on the beach and smoke pot from dawn till dusk.
from 10iii:
Those who have failed to adjust to the technical revolution facilitated by machine learning will have less to learn and less to do. There will become a class of knows and know-nots, replacing the haves and have-nots of the past. These classes will be populated by those who continue to invest in their personal education versus those who do not. Because of universal basic income everyone's survival will be assured. But the know-nots, lacking critical thinking skills, will become prey for political extremists, charlatans, religious cults, and schemes designed to extract from them even their basic minimum income.
from 10iv:
Know-nots, having been cheated out of the universal basic income, will also become have-nots. This will lead to crime and corruption in the inner city much as we see today, only with a more sophisticated high-tech flavor. Instead of gang members rolling in cars and committing drive-by shootings, they will hijack ubers and carry out their mischief using fleets of vehicles.
from 10v:
Transportation of people, packages and materiale will become hierarchical so as to exploit the economies of scale. Large modules will fragment into smaller modules for deployment on narrower roads and more specific arrival locations. Having accomplished their objective they will then recombine into larger units to complete the transportation cycle.
