The Progress of Artificial Intelligence and What This Means For Our Future
From possibly sentient AI to robots with citizenships.
Photo Credit: Becominghuman.ai
Did you know there is already an artificial intelligence robot with Saudi Arabian citizenship? After hearing this, some may wonder what the big deal is if Google’s AI is sentient.
In October 2019, the Saudi Arabian government awarded Sophia the artificial intelligence robot citizenship to its country as a way to try and stimulate AI development in the kingdom.
Although the press highlighted the situation’s hypocrisy due to how females in the country remain viewed as inferior to males, there was very little outcry about Sophia having consciousness. This citizenship gave Sophia entitlement to the laws of the country like humans, and there was a lack of discussion on where this might go. As a mater of fact, after a brief week of news blitz, this story pretty much died until it was revived last week when the documentary “Sophia” premiered on Showtime.
Photo Credit: TV Insider
So, what is the big deal about the Google senior engineer, Blake Lemoine, being suspended from Google for announcing to the world that the company’s artificially intelligent chatbot generator may be sentient?
Based on Lemoine’s interactions with the bot, he believes we have arrived at the point in civilization where AI processes are humanlike, especially now a desire to survive as an independent life form.
In addition to your worries about the war between Ukraine and Russia, recent supreme court announcements, climate change, the recurrence of COVID-19 and the increased chance of a global economic recession, do you have to put AI on your top worry list about the future?
Well, here are some facts to help you decide.
1. We have been teaching the importance of AI and why students and non-tech professionals need to at least grasp the fundamentals as to what is going on. This is an effort to ensure the world doesn’t end up with an evolved process smarter than us, but not exposed to much more than the bias of a country or company for over five years.
Right now, we see three AI processes with advancements worth following to make sure this is not happening as we speak. These include GPT-3 by OpenMind, owned in part by Microsoft and Elon Musk ventures, Google’s DeepMind and Wu Dao 2.0, launched by the Beijing Academy of Artificial Intelligence. We have talked about these three processes in our post from March 23 saying in the next 12-18 months lots is going to happen… and this is one example.
2. All three processes have equaled or surpassed what we call human intelligence in three out of the four areas needed to qualify as humanlike. The first area is memory. Think about this: would you rather have your memory or Google’s memory? The second area is rational thinking, in which a computer program functions much more rationally than any human thought or action process. The third area is learning. Machine/deep learning is quietly taking over all the AI applications coming to market.
The only area standing in the way of making AI processes able to interact and do simultaneous complex problem thinking is consciousness. This basically states the process has its own individual sentiment and desires. No one believed Sophia could do this, but thought it would be more like a throw back to you what it has learned, not learned itself. However, Lemoine states that this is now the case with Google.
3. What we do know about these languages, and perhaps the ones Facebook and Apple are working on, is that progress has exploded in creating ways for these processes to solve complex problems by themselves and creating new ways to think about reality especially in scientific areas.
For example, let’s take a look at how Google gave away AI that creates new human genetic code to fight diseases. More than 900 processes and drugs are being explored as we speak. Here is our June 10 post on this.
4. Finally, new theories are coming up to be used to make AI processes even more sentient, and the results are just starting. These include Andrew Ng’s DataCentric approach, which states we need to really make sure we are using the right data to help AI processes think, not just scale and use more data. Here is our May 2 post on that.
Not every area of AI processes need to equal or excel the best general intelligence approaches, rather special areas in AI can excel while others don’t. This is called human intelligence, not general, and this could lead to faster breakthroughs in certain areas. An example of this would be a world class nuclear AI scientist doesn’t have to be a world class musician.
So, as you can see, we have covered a lot of this in our newsletter for you and will be doing a podcast on it soon to get you ready for all of your “the world is crazy” conversations.
Once again, we can’t tell you whether we are at the point where AI processes are equivalent to humans, but we can state the following suggestions.
1. As Henry Kissinger and Eric Schmidt wrote about in their book “The Age of AI and Our Human Future,” our reality will soon be expanded from AI processes as well as ourselves and maybe even more fantastic than what we create. We need to set up policies and procedures on how to work with these processes from a history, philosophical, journalistic, legal and ethical basis… NOW!
2. The less open corporations are in sharing information, such as Google with DeepMind and the Chinese government with Wu Dao 2.0 versus what GPT-3 is doing, the more chance of mistakes not being corrected. This also goes for rules and regulations regarding development just like nuclear power, which need to be created and enforced.
3. Finally, we won’t totally leave this article without giving you three possible answers to the question we started with regarding AI. Is it at the point of sentiment and individual thought?
First, it is just mimicking what we know and telling you what it thinks you want to hear, so not really there yet.
Second, it is like Mr. Lemoine said, and acts like a young 7- or 8-year-old adult.
Or third, it is already way above us and creating a new order, we just aren’t aware this is happening, which would make the most sense if the process has evolved to super artificial intelligence like some believe will happen soon.
Stay reading as we try to keep you counted in these discussions, not out!
For Further Information:
Here is more information on the Google engineer who believes the company’s AI is sentient.
Learn more about Sophia’s documentary and the hypocrisy of her citizenship.
Here is a Henry Kissinger book review written by GPT-3.
Best,
Craig