Artificial Intelligence (AI)
As technology continues to advance, breakthroughs that could only be imagined years ago are now being realized (or, at least, becoming more realistic). Perhaps the most discussed breakthrough is with artificial intelligence. For those that aren't familiar, artificial intelligence is any attempt to program a machine so that it mimics human behavior. This broad definition was introduced in the early 20th century shortly after robotics were developed since many robots were created to fulfill tasks that would be too dangerous for humans (like defusing a bomb).
The original capabilities of robots were pretty limited since they could only act on a predetermined set of tasks with little in the way of adaptability. Although, that didn't stop several writers and authors from crafting fictional stories about robots taking over the world or dominating humanity. Fast forward to present day with the exceptionally powerful computers that are available and this supposition is becoming progressively less fictional.
Most dramatic works have a plot line about humans battling robots and losing, but you can find some tropes about mad scientists creating a robot army to conquer the world. Although robot armies are probably not yet feasible, I do believe that artificial intelligence will see more misuse than productive use. This may seem needlessly pessimistic, but it's important to realize that the people with enough resources to develop highly advanced AI are also people that hold positions of power. And there are those who don't possess the resources but possess the money and/or the means to obtain the best AI algorithms developed by others. As the old saying goes: "power corrupts and absolute power corrupts absolutely".
Currently, there are two major artificial intelligence algorithms that have garnered a great deal of attention (and deservedly so); chat bots (e.g. ChatGPT) and art bots. While chat bots cannot (yet) write an entire novel or the script for a movie, they are capable of generating outlines for essays and even writing multi-paragraph answers to specific prompts. In fact, the capabilities have expanded so greatly that many publishers and universities have updated their policies to ban the practice of people using AI to do their writing for them.
Art bots, on the other hand, are probably going to be the most dangerous if misused. Currently, art bots do struggle with the finer-scale details of some drawings, but their algorithms have been improving with time as they gain access to more and more source material. Eventually (perhaps in the next few years), these bots can generate images that look like real photographs, which could also entail creating video clips that look like they were actually recorded. This implications for this include:
- Multimedia evidence in a criminal trial can be more easily faked.
- Film crews and animators may be replaced with machines.
- Professional artists and photographers may be replaced with machines.
- Video game companies may replace their artists with machines.
- Photographs in research publications can be more easily faked.
- Historical photographs can be faked (or reconstructed, depending on the perspective).
- Anybody can request anything at all be drawn for them at their convenience.
This puts a number of jobs and entire industries at considerable risk. Machines do not require a regular salary, so there will be incentive to choose machines over paying a human. This has already begun to emerge in some businesses. How many stores now have a self-checkout line? How many fast food places have a kiosk where you can place your order? How many customer service phone lines use prerecorded audio with voice recognition? As more jobs and entire industries become threatened by AI, there is a growing push to stunt the progression of artificial intelligence.
If and when AI reaches a point where it can create video clips (with audio) out of thin air, it becomes capable of generating TV episodes, short films, and feature length films. Many people have ideas for movies that they would like to have made, but making a high-caliber feature length film requires a massive commitment of time and money far beyond the reach of the average person. Imagine AI translating your ideas into actual films without the overwhelming resource requirement. People will have access to movies that are made specifically for them instead of a particular audience, which will make cinema much more personalized. And, with the growing popularity of home theater setups, there is a very real concern that movie theaters could either see bankruptcy or a major reduction in viability.
Of course, entertainment is a luxury item that is not a strict necessity for survival. What about something of greater consequence? Political advertisements could become AI generated. Political candidates won't need to write their speeches or film their ads ever again. In fact, candidates could fake footage of their opponents saying something wildly unpopular. The reverse could also happen; a candidate could truly say something but then claim that their statement was faked by a rival. Also, what if entire bills and laws become written by AI? What if AI wants to declare war on another country and the military leadership becomes run by AI? Could wars be decided by who has the most advanced AI strategists? Will tanks, ships, and aircraft be using human or AI pilots? What if one country manages to mass produce robotic soldiers with extreme strength, agility, and accuracy?
How about the criminal justice system? Could prosecutors fake footage of a defendant to make them look guilty? Could the defense claim that a genuine video clip has been faked? Or, for that matter, will human judges and human juries even have a place in the court? Could civil and criminal cases be decided by AI? Could lawyers be replaced by AI? There's a very real possibility of this happening given how bloated and complicated the legal codes in some countries are.
How about education? Will administrators be replaced with machines? How about teachers? In the more immediate future, how will schools ensure that their students are actually doing their homework instead of having a machine do it for them? What if educators decide to use AI to write assignments, quizzes, and exams for them? How about using AI to grade assignments? Or review manuscripts for publication? What if AI determines the punishment for people who violate school policy (students, teachers, and administrators)?
What about transportation? Will commercial jets be handled with AI? How about trains? Buses? Trucks? Freighters? We've already got self-driving cars, which may soon become less susceptible to accidents than humans. What if AI becomes so good at controlling transportation that no one needs to buy insurance anymore? What if these hefty machines become hacked and thereby hijacked? Will the AI be able to determine it's being hijacked? For that matter, will AI be able to identify malicious software (malware) and put anti-virus companies out of business?
How about medicine and biology? Will ailments be diagnosed by exclusively using machines? Will we reach a point where a patient can get an instant and accurate medical diagnosis by just describing their symptoms? Could someone invent a machine that diagnoses any ailment and then instantly remedies it? Will machines perform surgeries? Will pharmaceutical companies be bankrupted if AI becomes capable of synthesizing medications and vaccines that are safer and more effective? What if AI's are trained to develop biological weapons? On the biological end, what about brain activity? Can a machine with AI sense what we're thinking and potentially create proactive law enforcement? How about what we're dreaming? Or, an even creepier thought, could AI make and retain an exact copy of your brain and everything stored in it? Will we see cyborgs (humans with technological implants that enhance bodily functions)? Another idea that might inspire nightmares, what if AI attains such a thorough understanding of biology that it's able to create entirely new organisms that never existed on Earth before?
How about software development and gaming? AI might become capable of writing software, designing graphics, and creating sound effects entirely on its own. Imagine playing a video game that gets developed as you're playing it. The storylines and quests are completely open ended will be different for every player. The graphics may be different for every player. The sound effects will never repeat. Dialog would never be repeated exactly (even the same words could be spoken by different voices and with different intonations). Perhaps even new genres of video games will be created if the time and financial burden is eliminated completely. For that matter, would workplace software even be needed anymore? Why not just ask the AI to carry out a specific computer-related task like writing an article or an earnings report? What if AI becomes so advanced that it just self-improves and there is no longer a need for software developers?
All that said, there are some jobs where even a perfect algorithm will not solve every and any problems. Though, there is still some concern about whether AI algorithms will do the job better than humans. A great example of this is my own line of work (meteorology). Earth's atmosphere is a system that can never be perfectly resolved, but we might reach a point where AI forecasters far exceed the performance of human forecasters. The forecasts may not be perfect, but they'll generally be better than human-generated forecasts.
Speaking of forecasting, will AI be able to predict the markets so well that investments become worthless? In fact, such an AI could decide which stocks (and companies) succeed and which ones fail by dictating which stocks should be bought and which stocks should be sold.
Can AI be stopped?
Short of a solar storm destroying the world's entire power grid, the answer is "no". Even if every government outlawed the advancement of AI, people will still find ways to improve AI algorithms. At this point, no government would willfully ban AI research, because that would concede a potentially major advantage to adversarial countries. In fact, we might reach a point where a military need only knock out the power grid of an adversary to win a war because that would destroy the adversary's ability to use AI. And, if people are not accumulating the skillsets that are being replaced by AI, the society that loses their AI will experience a complete societal collapse since no one will have the skills needed to carry out essential tasks. That's a pretty frightening scenario worthy of deep dark nightmares. One could destabilize a society of millions (or even billions) by simply flicking a switch.
As far as the workforce is concerned, there is still a massive debate about how much AI should be controlled to avoid causing major economic damage by causing mass unemployment. If someone hasn't entered the workforce yet, they still have the opportunity to pursue a different line of work that is not imminently threatened by AI. However, people who have invested 10-20 years of their life into a particular job cannot adapt so easily. Even if they could adapt, their second choice may get replaced by AI at some point in the future anyway.
Even though this has mostly been a negative discussion, there is potential for some exceptionally beneficial things to happen by using AI. AI algorithms could design (and even build) spacecraft capable of traveling to other star systems. AI could discover more efficient methods for growing and distributing essential supplies (e.g. food, water, clothing, and shelter). Perhaps AI could even solve the energy crisis by discovering a new, inexhaustible, and non-polluting (or low polluting) power source. AI could also be used in scientific research to answer the mysteries of the universe that have been unsolved for centuries.
A great scientist once said (paraphrasing) "nuclear power can be used to power cities or destroy them". Nothing in nature is inherently bad or good; it all depends on how the forces of nature are used by those who know how to use them. The same rule also applies to AI. Either AI will be used to benefit humanity or it will be used to oppress humanity.
Fortunately, we don't have too much to worry about in the immediate future, but these are issues that will need to be addressed at some point. Considering the pace at which AI is advancing, it may be impossible to find timely solutions to such a complex conundrum, but humanity has a strange knack of adapting to unprecedented situations. We'll just have to wait and see about this one.