How AI Is Saving Humanity
"The first ultraintelligent machine is the last invention that man need ever make." This statement was made by mathematician Irving John Good in 1965. He was envisioning a machine smarter than any human who had ever lived, one that would design even smarter machines and leave humans in the dust.
Now, while we still haven't created an ultra-intelligent machine, we have successfully created something that could end our species. But if used correctly, could also save us.
In the history of our species, we have been remarkably skilled at inventing and using tools to further our civilization. From the stone axes and spears of our ancestors to steam engines and computers, the knowledge and intuition used to create these tools has allowed us to improve the quality of our lives tremendously.
Today, we stand on the precipice of a new invention, artificial intelligence (AI), the next chapter in our story. But unlike the tools of the past, AI could do both harm and good. So what if we invented the wheel but didn't know how to use it? So what if the light bulb was never imagined? If these inventions failed, the most likely outcome is that our civilization would probably just continue the status quo. But with the addition of AI, things are about to change.
On the one hand, a sentient AI, if that's even a possibility, could dethrone humans as the smartest species on Earth and try to take over the planet for its own benefit. But if we can harness the power of artificial intelligence and put it to good use, it could potentially save us and the entire planet. This is how AI will save humanity.
AI has advanced rapidly in recent years, which is why visions of sentient machines taking over the world have been dominating the news cycle. There is, and rightfully so, a lot of criticism surrounding the rapid development of artificial intelligence. We've made several videos talking about the dangers of algorithms and AI tools like ChatGPT right here on this channel. But among all of that, there are a lot of positives that have come with the development of AI, some of which are already revolutionizing our world.
Cancer is one of the biggest hurdles we have to face as a species. Research shows that if you live long enough, cancer will eventually kill you - if you don't die of something else first. 1 in 2 people in the world will develop some form of cancer during their lifetime. The numbers are scary, but they might not be for much longer.
Artificial intelligence is helping to advance cancer treatment. By quickly understanding how cancerous cells become resistant to anticancer drugs, AI tools can help to massively improve cancer drug development and use. Pharmaceutical companies are using AI to scan through large volumes of data and use predictive analysis to figure out which molecules are best suited for use in medication to fight cancer. And it's not just theory.
In a recent study, researchers from the University of Toronto and Insilico Medicine used a computer program called AlphaFold along with a tool called Pharma.AI to find a new way to treat liver cancer.
The AI tool found a new target to attack the cancer and also found a molecule that would stick to that target. This molecule could be included in a new cancer treatment drug. The researchers completed all of this in just 30 days. Imagine what they could do with more time and more powerful AI tools.
Artificial intelligence is also being used in medical imaging. Analyzing CT scans, X-rays, and MRIs to find lesions or other abnormalities a human radiologist might miss. These are pattern-oriented repetitive tasks – exactly what machines excel at.
Even if AI is not able to assist with critical areas like surgery or specialized care just yet, if it can improve the productivity of medical professionals by 2-3 times, which is probably a conservative estimate. It might just be that we have a healthcare revolution on our hands.
Research shows that by the year 2034, there may be a shortage of up to 48,000 primary care physicians. Tools like these might allow us to bridge the gap between the amount of care we require and the number of physicians available to give us that treatment.
There was a recent incident where an unknown tick-borne disease on a dog was producing confusing symptoms. The dog’s worried owner put the details of its symptoms into GPT-4, which hypothesized what the condition might be. The owner took this information to a second veterinarian, who confirmed one of the probable diagnoses that GPT-4 had suggested. While the puppy still definitely needed to see a real vet, GPT-4 was able to massively speed up the time it took to diagnose the illness. Today, the dog has made a full recovery, thanks in part to GPT-4.
AI can also assist people living with disabilities enabling them to live more independently. GPT-4 is being incorporated into apps like "Be My Eyes" and "Virtual Volunteer" to help the blind and visually impaired to better interpret the world around them.
We also now have nearly accurate real-time captioning software that allows people with a hearing impairment to watch movies, follow along with online classes, or even take calls from loved ones.
AI has the potential to create life-changing opportunities for people living with disabilities. It makes it easier to create interactive tools to support both physical and mental accessibility and to promote independence.
Speaking of mental accessibility, mental health issues have been on the rise in recent decades, placing a significant burden on individuals, families, and society as a whole.
AI can be used to assist the creation of diagnostic tools, personalized treatment plans and even provide virtual therapy through chatbots and other interactive platforms. In fact, this has been a surprising reason a lot of people have been using ChatGPT lately. It's no wonder there was a significant drop in the number of posts per day on the relationship_advice subreddit right after ChatGPT's release. The immediate access, the complete lack of judgment, and its creative potential make ChatGPT an excellent mental health aid. It can help address the shortage of mental health professionals, increase access to care, and reduce any stigma associated with seeking help.
Education is another area where the powers of AI could be harnessed to do amazing things. People who are dyslexic have been flocking to Reddit communities to say how ChatGPT has allowed them to learn things at their own pace - something a traditional classroom setting could never provide at scale – and how they wished it had existed before.
We've had online classes before, yes. But though they were accessible, the content was never tailored to each person's individual needs. With artificial intelligence tools, you can create that with just one prompt.
Of course, there is a trade-off here. Many students have simply started copying and pasting information given to them by AI without actually reading or understanding any of it. People are genuinely worried that this might cause students to lose interest in learning anything. Why bother when they can just ask ChatGPT to spit out the answers to their assignments? But is this the fault of the tool or of our current education system?
Let's consider a similar scenario. One of the greatest capabilities of ChatGPT is writing and debugging code. You might imagine this would discourage people from learning to code. That is until you read about people who have, for the first time in their lives, found a friend, so to speak, who will not only give them examples of good code. ChatGPT can also tell them what mistakes they made and speak to them with a respectful tone, as opposed to coding forums that are known to criticize users for asking "too obvious" questions.
The potential of AI in education is huge, with its ability to customize learning experiences to individual students and to bridge the gap between well-resourced and under-resourced schools. By identifying and addressing each student's unique needs, strengths, and weaknesses, AI can encourage a more inclusive and effective learning environment which has the potential to reduce educational inequality.
One of the GPT-4 demos included writing the code of a website from a rough drawing on a napkin. Imagine how much power that gives a small business owner to start their own project. Something that previously would have required a lot of money and time can now be done with a few well-written prompts.
There is an obvious concern about job displacement with all of this. But how many people live to write emails? How much meaning does one get by spending hours debugging code only to find out what was missing was a semicolon?
Wouldn't we rather spend our time on more meaningful pursuits? Trying to understand the meaning of life and our place in the Universe? These are the areas that Large Language Models are not able to compete with humans. And without a fundamental restructuring of their architecture, cognitive scientist and AI researcher Ben Goertzel (gort-zl) thinks they are never realistically going to be able to think like that anyways.
Purely from a knowledge and research perspective, even though AI isn't intelligent enough to make decisions on its own, just being able to summarize large quantities of information will massively assist innovation and research. Combined with its teaching abilities, power to analyze large quantities of data, and ability to brainstorm, you have an information juggernaut on your hands that will revolutionize the way you learn and understand things.
This even applies to the wisdom of the past. AI can help with the preservation and dissemination of human knowledge and cultural heritage. As our world becomes increasingly digital, there is a risk that important historical artifacts, documents, and works of art may be lost or forgotten.
AI can assist in the digitization, organization, and analysis of vast amounts of cultural data, ensuring that future generations can learn from and appreciate the accomplishments of those who came before.
ChatGPT is fundamentally a language model and has now been used to speak languages that are nearly extinct. This is absolutely vital to their preservation. In fact, ChatGPT was recently used to re-create native-sounding phrases from the Chinook Jargon Language, a Native American language that’s almost extinct.
Now, take a moment to imagine harnessing all these powers to solve the most dire problems that our civilization faces.
Whether it's climate change, an asteroid impact, another raging pandemic, or depleting energy resources - artificial intelligence can help with ALL of these.
It can legitimately accelerate innovation. Programmers can be more efficient. Researchers can churn out more output, and the healthcare system can ease the pressure and be prepared for when it is really needed.
I mean, looking at all of this, isn't it immoral not to embrace AI at this point?
"What inventions might a superhumanly capable Artificial General Intelligence make?" asks Ben Goertzel, referring to a machine similar to the one John Irving Good also imagined. "Perhaps little things like curing cancer, death, and mental illness, solving climate change, space travel, mind-uploading, cheap food, fusion energy … an era of abundance in which nobody has to work for a living, and people can focus on social, spiritual, artistic and intellectual fulfillment. …"
Or as AGI researcher Joscha (Jor-sha) Bach put it: "There may be a 10% percent probability that people will die if we build Artificial General Intelligence, but there is a 100% probability that people will die if we don't. Especially you."