An intro to emerging tech in AI, and why we might want to prepare ourselves for change ahead of time
As humans, we’ve generally agreed and adapted to the fact that computers are better than us at chess, computation, and storing knowledge. But not everyone has mentally prepared for just how much AI will impact all of our lives in the near future. There are many opinions about how AI will affect the world in the next ten years, but it is clear that we are about to enter an era where machines will be able to do things that were once thought only possible with human intelligence.
The question is, will this be a good thing? Will AI help us or hurt us? And what do we need to do to make sure that AI benefits humanity as much as possible? The truth is, no one knows exactly how AI will affect the world. Some people say that AI will be harmful because it could take away jobs from humans and put them out of work. But others argue that this is a misconception, since many jobs that are currently done by humans are actually repetitive and boring. Machines could do these jobs more efficiently than people, which would give us all more time to focus on things we really enjoy doing like art or music. Then again, new forms of generative artificial intelligence are already creating art and music that is perhaps objectively better than most of us humans could create anyway. As AI becomes more sophisticated and capable of making decisions that affect our lives, we will need to ask ourselves: who is responsible for the actions of machines? Will governments be able to regulate them? And if so, how far should regulation go? How much control should people have over their own data, which is often used by companies like Google and Facebook to develop AI systems?
Let me pause here to describe a theme that I would like to discuss in this post:
I believe we need to prepare ourselves individually and collectively for deep socioeconomic change brought on by innovation in artificial intelligence and machine learning. Our lives will be impacted faster than we think, and we will likely need to reconsider how our routines, careers, and economies are organized.
- Already, AI is generating content, driving cars, analyzing data, and more.
- Very shortly, AI will be reviewing legal documents and giving legal advice. It will give medical advice, assist with surgery, perform financial analysis, do your taxes, build your car, and more.
- In the next several years, smart machines will be the majority of the workforce in production environments.
- Since humans make up the major cost of most goods and services, outsourcing that labour to machines will create massive amounts of wealth as the cost to produce goods and services will approach $0.
- The pace of scientific discovery will accelerate rapidly as highly efficient programs will make discoveries on our behalf.
- Space exploration technology will advance rapidly and become much cheaper as very smart machines will realize they need to extract more resources to accomplish their goals, and will hopefully be smart enough not to extract them all from the earth.
Some of this may sound like it’s very far away considering the idiotic so called “artificial intelligence” examples we run into today. Like being locked in your Tesla with no way out while it bursts into flames for example. But consider Moore’s law, which said that roughly every two years, the number of transistors we can put on a microchip will double. This has been happening since the 60’s. From what I can tell, we seem to be experiencing a similar law with advances in machine learning and intelligence.
There are thousands more people working on new use cases and innovation in the industry, and more join every day. The tools are also becoming less complicated and therefore more people can use them and contribute to them than just ultra smart linear algebra nerds.
Couple that with the fact that we have figured out how to train computers to learn for themselves, which means machines with infinitely more processing power than humans, who don’t need to sleep or take care of families, can work around the clock on solving problems for us.
Mathematician John Von Neumann, coined a term called “The Singularity” which is a concept that at some point, technology will progress so rapidly that it will become out of control and irreversible. In an artificial intelligence context, this could be when an AI becomes smart enough to teach itself, at which point it could rapidly consume the internet, and become infinitely smarter in a short amount of time. At that point, an infinitely smart AI which was designed by humans to achieve a goal, may make unexpected and possibly dangerous decisions in order to accomplish that goal. (A great place to better understand the dangers of AI becoming extremely intelligent and what we can do to manage them safely would be with a book called Superintelligence).
This is why I think we are approaching a no turning back point for artificial intelligence. I think it will very shortly be smart and efficient enough to exponentially grow its reach and impact in our lives, careers, and economies. And if we don’t pause and consider what that could mean for us, we will be caught unprepared when it all happens.
To illustrate this new reality, I’d like to share some of the innovations emerging under a new category called generative AI. Once we have a context for what generative AI is, and how it may expand, we will be better prepared to ask questions about how we may be impacted, and what we could do to adapt to these changes so they are positive for us.
Generative AI is a field of machine/deep learning that utilizes massive neural networks of information to create things that are entirely new.
Machine learning is the process by which we train computers to do/learn things in a structured environment. That is to say, you tell it the right way to do things, and give it the right answers, and over time it learns and can do it without help. Deep learning however, is when you give a computer a massive amount of data, and then allow it to process the data itself with no supervision. The computer will search for patterns and relationships that would be impossible for humans to detect considering how large the data set is. But since it has the ability to analyze huge amounts of data quickly, it can actually learn with no guidance from humans.
Let’s look at some examples of Generative AI.
Examples of generative AI in use today
Of the exactly 300 words of writing in the first two paragraphs of this post, my brain wrote the first sentence consisting of 22 words. The rest was written in seconds by a generative AI writing platform. I chose the tone of the writing to be “persuasive” as opposed to cheerful, or relaxed, or bold, or several other options, then I primed the tool with the first sentence and it was on its way. I chose to cut it off so I could start writing in my own words again, but it could have carried on for pages.
The featured image of this post (which I will display again below) was generated artificially as well using a tool called Dall-E. Dall-E is software that allows you to provide a written prompt and receive a completely unique and artificially generated photo output.
I provided the following prompt for the featured image of this post:
Baroque style painting of 16th century farmers working in a field with robots
Here are a couple other random pictures I generated and the prompts I used. There are quite a few, but I couldn’t help myself, it is just too fun to use. You should consider trying it yourself, it is addicting. (Dall-E).
Surrealist painting of people surfing, but they are in space
A classy cologne advertisement, but the model is sick
Chris Farley flying with a jetpack over the city of Vancouver
Portrait of a hamster wearing royal robes and a crown against a royal purple backdrop
A realistic pencil sketch of a juicy hamburger
A retro photograph of a man standing in his 1970’s style home. He is wearing thick rimmed classes and has a very large smile
A cartoon of a man driving a tractor on the highway and holding up traffic
If you feel amazed by this, you are not alone. However, the generative abilities of AI go way beyond simple text and image generation.
Check out podcast.ai. On its home page, you will see a couple podcasts that were generated by an artificial intelligence. The site generated a fair bit of press for releasing a podcast interview between Joe Rogan, and the late Apple CEO, Steve Jobs. The voices sound incredibly authentic to Rogan and Jobs and the conversation flows naturally and follows subjects that make sense given the context of Steve Jobs being interviewed.
Here are some other applications for generative AI that are happening now that I found through a work colleague.
Take a couple still photos with your phone, and AI will generate drone image from it.
Place your companies products on any artificially created backdrop to create great marketing assets
Generate music from text
Create a visual representation of what you are reading
Generate fashion
A bunch more ideas
ChatGPT
I have to zoom in on GPT-3 (ChatGPT). It’s getting a lot of attention right now. It was a lot lesser known even a couple weeks ago then it is now. I had no troubles logging in and using the service last week. When I log in now, I get messages that they have high demand and are working to scale systems. I wasn’t even able to login in some cases due to excess usage.
ChatGPT has big implications on the way we browse the web today. It may be a chink in the armour of Google, which up until now, its language model and search engine has been an unstoppable force. Think about your experience on Google today. You search for something you want to know, then have to scroll through 3-4 ads to get to the real results. Then you have to scan the results and see what page will most likely offer you the information you need. Clicking on the websites you think has the info you want means you have to accept cookies, ignore a popup, and close the giant video ad that just opened. So annoying!
ChatGPT can remember what you said/asked earlier in the conversation. It can allow for follow up corrections, challenge assumptions, and even admit when it was wrong. Best of all, there is no “browsing” required for your answer.
I couldn’t believe how detailed some of the answers I could get when prompted. For example, I asked it to give me at least a 2000 word explanation of how distance is measured in space (which I previously wrote about here, so I had an idea what I was expecting it to say). It gave me a bunch of detailed writing, and even provided nicely formatted formula’s for how to calculate distances.
Turns out GPT-3 can do even more than I was expecting. Here is a video of someone asking it to write a custom plugin for WordPress, and the program instantaneously writes all the code for it.
Maybe you are applying for a job and need to write a cover letter? Ask the program to write you a cover letter for X position, and address it to X company and watch your calendar burst open with new time!
Kids, if you are reading this, you should not use ChatGPT to write your book reports or do your homework for you.
ChatGPT is an example of using deep learning to program a computer. So when you ask the chat bot a question, it doesn’t actually know the answer. All it knows is the relationship between the words, sentence, and idea that you asked it, and it provides you back a human sounding response based on patterns it analyzed from huge amounts of written content. This does mean that the bot may reply to you with a convincing answer, that is actually very wrong.
If you want to keep up to date with the huge amount of innovation happening in this field you could follow this site called Bleedingedge.ai.
Or check out this really nicely presented post by Sequoia Capital about Generative AI. It does a nice job at laying out the landscape and timelines of AI.
A lot of the examples I showed above existed in the virtual space. So content creators, marketers, designers, videographers and other related roles are likely on the front line of disruption. But I would return to my point about the exponential growth potential of computers that can learn for themselves, and it shouldn’t be hard to see how the disruption spreads broadly.
How should we prepare? What should we think about?
Ideally, everyone should be able allowed to participate in the gains and benefits driven by automation. The automation of goods and services should drive the price of goods and services way down. This means that everyone becomes wealthier, because their buying power has just gone up relative to the price of goods. A great article called “Moore’s law for everything” speaks to this. The idea is that we should expect a similar “Moore’s law” phenomenon to happen to the price of everything, except in a downward direction. Imagine as AI becomes better and better at building homes, delivering education, making food, that the price of these goods halves every two years. This gives me a lot of hope when it seems like the only trend for goods/services is to get more expensive. But Moore’s law for everything doesn’t read like fantasy. If humans are rarely needed to build a house, and technology continues to improve at its typical pace, the price to build a house will become cheaper and cheaper over time as efficiency grows. Imagine building a house, and the only cost is the materials. The constraint in the house example is resources like concrete, but that is another reason why space exploration will become prominent. The resources will be extracted from places other than earth.
We shouldn’t think about robots taking over all the work we are familiar with today as meaning we will have nothing to do. We will discover new jobs, we always have after technological revolutions. The jobs will be a lot more creative than they are now though, as we are pulled out of the drudgery and routine of tasks that simply need to be done.
There is no question though that our working routines and environments will change. I don’t think 20th century notions of education, policy, work, and economies will be fit to handle the changes that are coming, and we will require a willingness to make large changes to some of these systems that are deeply ingrained in our lives. Big changes always result in discomfort and pushback from people. These tensions will have to be managed though, as we can’t afford to ignore these issues. If we do ignore it, my worry is that this technology will become far too centralized, and wealth inequality will spiral out of control. Not only that, we may just extinct ourselves by opening the floodgates and letting loose completely unbounded super-machines into the world.
Let’s dig into a few major pillars of our societies that will likely need changes.
Education
As AI automates more tasks and jobs, workers will need to acquire new skills and knowledge to remain employable. Humans will likely need to focus on a combination of skills that would be difficult for AI to replicate, like critical thinking, creativity, collaboration, problem solving, and emotional intelligence.
There are offshoots of education that will be required for this new reality as well. AI will create many social and ethical dilemmas that we need smart and well balanced people to address. These people should understand how AI works, thinks, and be able to propose ways to address the social and ethical issues we will encounter.
We probably also need to build in general education of AI into our schools. I think it will be important that as many of us as possible understand the opportunities, risks, limitations and biases that come with AI. Given how powerful these computers/machines/robots will become, we should all have a general understanding of how to minimize risk. Because this technology is so new, and unprecedented, most of us won’t even comprehend the kind of damage we may carelessly inflict on people and the earth by attempting to do something seemingly benign with AI.
For example, the book Superintelligence uses an example of an extremely smart AI that was programmed to manufacture paperclips. If we just let an extremely smart computer go off and design paperclips, it may destroy the earth in the process by extracting every possible resource in order to create paperclips. So maybe we assign a limit, like “make only one million paperclips”. But the author again said that an extremely smart computer would likely not ever assign a 100% probability to the fact that it completed its tasks, and may still keep making paperclips to be very sure that the probability that it truly made one million paperclips is as close to 100% as possible.
Work
We may need to redefine the role of work in our societies, and come up with creative ways to value human effort.
If the labour associated with producing basic necessities like food and shelter is outsourced to machines, humans may shift to focus on things other then the basics. Perhaps jobs tailored to deeper human needs like relationships, art, music, fitness/nutrition, connection, faiths, etc become higher prioritized.
It will be interesting to see how we define and attach value to jobs in the future. Those working directly with AI to produce transportation are surely valuable, and there are tangible items to attach to their output. But how do you measure the value behind adding meaningful connection to a persons life? Surely working against the loneliness epidemic many of our cities face should be assigned value? We don’t want to incentivize people to only work directly with AI to produce material outputs. We need a well balanced society that has the interests of humans at heart, not just consumption or technological improvement.
Policy
AI is a complex and rapidly evolving technology, and existing policies and regulations are not going to be well-suited to address its challenges and opportunities.
How we incentivize people under capitalism, and how we redistribute income through taxation may need adjustment. Income tax for example, will become an increasingly smaller annual revenue stream for our governments, as the amount of income people earn will diminish as they perform less labour. So how is income redistributed then?
I’ve read arguments that we should change our tax systems from taxing labour, to taxing capital. This sounds pretty radical in a capitalist system, and may have people worried that it will decrease innovation and investment in capital/technology, which is how our societies advance. I think that is true in todays context of labour and capital, but I am trying to remember that the dynamics between the two will shift greatly as AI takes over. As we introduce more and more automation and intelligence into the world, our ability to create new technologies, innovate, build, and research will increase exponentially. I am going to paste in a section. of writing from the Moore’s law article as it offers a “conversation starter” idea for two new ways to tax.
We should therefore focus on taxing capital rather than labor, and we should use these taxes as an opportunity to directly distribute ownership and wealth to citizens. In other words, the best way to improve capitalism is to enable everyone to benefit from it directly as an equity owner. This is not a new idea, but it will be newly feasible as AI grows more powerful, because there will be dramatically more wealth to go around. The two dominant sources of wealth will be 1) companies, particularly ones that make use of AI, and 2) land, which has a fixed supply.
There are many ways to implement these two taxes, and many thoughts about what to do with them. Over a long period of time, perhaps most other taxes could be eliminated. What follows is an idea in the spirit of a conversation starter.
We could do something called the American Equity Fund. The American Equity Fund would be capitalized by taxing companies above a certain valuation 2.5% of their market value each year, payable in shares transferred to the fund, and by taxing 2.5% of the value of all privately-held land, payable in dollars.
All citizens over 18 would get an annual distribution, in dollars and company shares, into their accounts. People would be entrusted to use the money however they needed or wanted—for better education, healthcare, housing, starting a company, whatever. Rising costs in government-funded industries would face real pressure as more people chose their own services in a competitive marketplace.
As long as the country keeps doing better, every citizen would get more money from the Fund every year (on average; there will still be economic cycles). Every citizen would therefore increasingly partake of the freedoms, powers, autonomies, and opportunities that come with economic self-determination. Poverty would be greatly reduced and many more people would have a shot at the life they want.
A tax payable in company shares will align incentives between companies, investors, and citizens, whereas a tax on profits does not–incentives are superpowers, and this is a critical difference. Corporate profits can be disguised or deferred or offshored, and are often disconnected from share price. But everyone who owns a share in Amazon wants the share price to rise. As people’s individual assets rise in tandem with the country’s, they have a literal stake in seeing their country do well.
Henry George, an American political economist, proposed the idea of a land-value tax in the late 1800s. The concept is widely supported by economists. The value of land appreciates because of the work society does around it: the network effects of the companies operating around a piece of land, the public transportation that makes it accessible, and the nearby restaurants, coffeeshops, and access to nature that makes it desirable. Because the landowner didn’t do all that work, it’s fair for that value to be shared with the larger society that did.
If everyone owns a slice of American value creation, everyone will want America to do better: collective equity in innovation and in the success of the country will align our incentives. The new social contract will be a floor for everyone in exchange for a ceiling for no one, and a shared belief that technology can and must deliver a virtuous circle of societal wealth. (We will continue to need strong leadership from our government to make sure that the desire for stock prices to go up remains balanced with protecting the environment, human rights, etc.)
In a world where everyone benefits from capitalism as an owner, the collective focus will be on making the world “more good” instead of “less bad.” These approaches are more different than they seem, and society does much better when it focuses on the former. Simply put, more good means optimizing for making the pie as large as possible, and less bad means dividing the pie up as fairly as possible. Both can increase people’s standard of living once, but continuous growth only happens when the pie grows.
I thought his ideas were interesting and they reminded me a bit of some of the discussion in The Future of Capitalism. The book also talked about the growing divide between cities and suburbs, and how those living in cities have profited greatly from the network impacts of cities drawing people in, and raising property values without really contributing anything.. Another major theme in the book was the idea that we can’t allow political extremes to twist things that are actually useful. The author talks about a shared identity, and love for one’s country for example, which is twisted by extremists to be nationalistic. But like the Moore’s law article states, it would be a powerful force to have an entire country invested and focused on seeing their country succeed.
I worry that most of our political parties have become too ideological to tackle these issues with changing policy, and that all we would get out of trying to constructively bring these issues up is flame wars. I don’t know how to get around this other than tackling it privately first, like a science experiment. We work and collaborate with others on ideas and solutions, develop and test hypothesis, and try to get public support once viable solutions are proposed. We’ll need a collective understanding that not everything will work perfectly. Our societies are too complex to simply be “figured out” and have a formula applied to them. Trying out a well thought out proposal, seeing where it went wrong, and correcting it is better in my mind than ignoring it, and engaging in political infighting for five years, and then realizing its too late.
In Summary
I hope this post was a good summary and introduction to some of the emerging tech coming out of AI at the moment, and I hope it spurred on some thinking about how our futures may be impacted and what we can do to prepare. I’ve barely scratched the surface with this post, but I’ve already spent a few weeks drafting this, and every day of delay makes some of what I’ve written more irrelevant, as the tech is progressing so quickly. So I feel a sense of urgency to post, and can perhaps follow up in the future.
I’d love to hear more people talking about AI and thinking through its implications, and am curious if/where there are discussions groups happening on the subject, from both specialists in the field, and the rest of us who will have our lives impacted at some point as well. If you are reading this and have any general thoughts, or thoughts on how your future might be impacted, I’d love to hear it, please leave a comment. From what I’ve seen, there is a lot of excitement and fear on this topic, both of which are very understandable.
We are witnessing the beginning of a modern industrial revolution, and it is happening whether we like it or not. If we are brave enough to embrace this new technological force, and if we adapt our societies to make it as accessible and advantageous as possible, the benefits to humanity are seemingly endless.