CAVEAT: The aim of this blog article is not to present a comprehensive guide to artificial intelligence (AI), but rather to provide parents and educators with a general understanding of the transformative potential of AI, and its associated challenges. We hope to equip parents and educators with the needed information to initiate conversations with their children or students about the positive and negative impacts of AI that are both known and unknown. We believe that these conversations on this specific topic need to start taking place both and home and in our schools.
Artificial intelligence (AI) has come a long way in the past decade. With significant breakthroughs in machine learning and deep learning over the past year, we are now seeing AI that can-do things that were once thought impossible, or only found in movies or on TV shows like Star Trek. As a result, many of the social media platforms that are most popular with youth are quickly integrating the newest form of AI, Generative AI, into their platforms.
There are two general types of artificial intelligence that parents and educators should be aware of:
#1 Discriminative AI (oldest)
Discriminative AI is designed to analyze patterns and make predictions. Examples of Discriminative AI include Siri, GPS, natural language processing, and medical diagnostic tools. These systems rely on predefined data and are trained to recognize specific patterns. Discriminative AI is great at performing tasks with a high degree of accuracy, but they lack creativity and the ability to create something new. Here are some examples of how Discriminative AI has been integrated into the onlife world:
- Image recognition: Discriminative AI is widely used in image recognition, where it can accurately identify and classify objects in images, such as faces, animals, and vehicles.
- Fraud detection: Banks and financial institutions use discriminative AI to detect fraudulent transactions by identifying patterns in customer data that indicate suspicious behavior.
- Language translation: Discriminative AI is used to translate text from one language to another, by analyzing and classifying patterns in the language and syntax of both languages.
- Voice recognition: Virtual assistants like Siri and Alexa use discriminative AI to recognize and interpret spoken commands and respond appropriately.
- Medical diagnosis: Discriminative AI is used in medical diagnosis to analyze patient data, identify patterns, and classify diseases or conditions, which can assist doctors in making more accurate diagnoses and treatment plans.
#2: Generative AI (newest)
In contrast to Discriminative AI, Generative AI has the ability to produce completely novel and distinctive creations independently, such as text, images, or music, that are imaginative and unforeseeable in their content generation. These creations are not based on pre-existing templates or patterns, but are generated entirely from scratch.
One of the most well-known examples of generative AI is ChatGPT (1),which can generate documents or answer questions based on word prompts. An example of a word prompt:
“Write me a high school level essay on the book Catcher In The Rye as if it was written by a high school student”
Within seconds ChatGPT will write a grammatically correct essay that can be handed in as a school assignment.
Another well know Generative AI is Dall-E (2), which can also create images from scratch using word prompts rather than traditional coding. An example of a word prompt in Dall-E:
“Create a picture of a high school student in a classroom sitting at a desk using ChatGPT to write an essay on Catcher In The Rye”
Again, Dall-E will create such a picture within seconds.
One of the reasons why Generative AI has come so far so quickly – its unique ability to learn and improve at a much faster pace than humans. As an example, to demonstrate the generative power of ChatGPT:
- In a USA S.A.T. test – ChatGPT scored 94%, the human average is around 50%
- In a USA law school BAR exam – ChatGPT scored in the 90thpercentile
- In a verbal-linguistic IQ intelligence test where a score of 140 is considered “genius” – ChatGPT scored 152, which placed it in the 99th percentile
To demonstrate Generative AI’s explosive adoption by the masses – ChatGPT has achieved the milestone of gaining 1 million users faster than any other publicly available technology in human history. As an example:
- Netflix took 3.5 years,
- Twitter took 2 years,
- Facebook took ten months,
- Instagram took 2.5 months,
- ChatGPT took 5 days
So, what is the difference between ‘Human Intelligence” and “AI Machine intelligence”?
Human Intelligence:
Human intelligence involves four interconnected human attributes:
- Intelligence (the ability to solve problems)
- Morals (the ability identify good vs bad)
- Consciousness (the ability to feel & experience pain, joy, love, anger)
- Survival Instinct (sex, hunger, sleep) that were key to human survival based upon evolution
AI Machine Intelligence:
AI machine intelligence lacks morals, consciousness, and instincts, thus necessitating human intervention to make these distinctions. As a Star Trek fan, Commander Data comes to mind – Data looked and sounded like a human in appearance and speech, yet acknowledged his lack of human intelligence, which he constantly sought to acquire.
Given that AI machine learning has no morals, consciousness, or instincts it can’t differentiate right from wrong, and therefore can perpetuate biases, spread misinformation, pose potential risks to privacy, and even be used for crime (3). The exponential growth of AI development and its potential impact on humanity is truly staggering. Even its developers are uncertain about its limits, which is a cause for concern shared by many, including us.
There are current examples of how GPT chatbots have produced offensive and harmful content in the past, including racist, sexist, and homophobic messages. In one case, it even played a part in someone taking their life by suicide (4).
There is a real risk that generative AI could be weaponized and used to spread misinformation or manipulate public opinion, as it has the ability to generate plausible-sounding content without necessarily being accurate or factual. This can be especially problematic in areas such as politics or public health, where misinformation can have serious consequences. Cambridge Analytica comes to mind (5)
There is also a potential risk to privacy posed by social media companies who are moving extremely quickly to integrate Generative AI, like ChatGPT, into their platforms – Snapchat’s “MyAI” is a good example of this. It is still unclear if Generative AI will collect and store large amounts of personal data from users. This data could be used for a variety of purposes, such as targeted advertising or even more nefarious activities such as identity theft or surveillance. This is one reason why the Canadian Privacy Commissioner has launched an investigation into the privacy concerns associated with generative Al (6).
Very recently, our friend and colleague Dr Sameer Hindujua wrote just an excellent article on some other concerns surrounding the misuse of AI that he titled “Generative AI as a Vector for Harassment and Harm” (7) Please have a read of Dr Hinduja’s article, a link to which you can find in the below noted references.
So, What Do We Think:
The emergence of Generative AI has been met with both excitement and concern from different stakeholders. Some proponents see this technology as a major leap forward in the development of AI, with the potential to improve various aspects of society, including education, healthcare, and business, which we think it will. There are those who believe that these AI platforms could transform how we communicate and interact with technology, making it easier for people to access information, communicate with others, and accomplish complex tasks that can’t be accomplished by humans alone.
For example, Generative AI has made it possible for people to interact with computers and other digital devices more naturally, using speech or text-based inputs. This could enable greater accessibility for people with disabilities, as well as improving the user experience for a wide range of applications. (8)
On the other hand, critics of Generative AI have raised concerns about the potential for unintended consequences and the negative impacts it may produce. One major concern is the potential for these AI programs to amplify existing biases and inequalities in society, especially if they are trained on biased or unrepresentative data. This could result in discriminatory or harmful outputs that perpetuate or even worsen existing social problems (9).
There are also concerns about the concentration of power and resources among a few large tech companies that are developing and deploying Generative AI. Some worry that this could lead to a lack of transparency and accountability in the development and deployment of large Generative AI programs, with potentially negative impacts on privacy, security, and human rights (10).
There are also concerns about the potential impact of Generative AI on jobs and employment. As these Generative AI programs become increasingly sophisticated and capable, there is a risk that they could automate a wide range of tasks, which are presently being done by humans, potentially displacing large numbers of workers and exacerbating existing economic and social inequalities (11).
While the emergence of Generative AI represents a significant and potentially transformative development in the field of AI, it also raises important ethical and societal questions that we will all need to address as we continue to develop and deploy these transformational technologies.
Well-known American inventor, futurist, author, and computer scientist Ray Kurzwell stated:
“An invention has to make sense in the world in which it is finished, not the world in which it is started.” (12)
This quote is particularly true for Generative AI. While we have made incredible strides in AI, we must also ensure that it is used ethically and responsibly. AI has the potential to revolutionize the world, but we must ensure that it is for the betterment of humanity and not its detriment.
The fast-paced incorporation of Generative AI into social media platforms has become a concern for us, due to the numerous challenges that are still unknown surrounding it’s use. Many social media vendors are hastily trying to integrate Generative AI it into their products, which has resulted in users, particularly teenagers, being used as experimental subjects in this novel and concerning social media experiment (13).
As advocates of the balanced and appropriate use of technology, we have concerns about the current AI arms race that seeks to monetize and gain market share, particularly when it involves young people, and their use of Generative AI technology on these social media platforms. We do think there needs to be government regulatory oversight to ensure that:
#1 – We, and our kids, are not used as the Guinee pigs in the development of Generative AI to the financial benefit of these companies, and
#2 – Vendors of Generative AI development must be regulated to the best of our abilities to mitigate the potential Pandora Box type risks associated with this technology’s development and its use.
Parents – We hope this blog article provided you with a little better insight into the current state of artificial intelligence, which you can now utilize to initiate meaningful discussions with your child.
Teachers – use this blog article to create a thought experiment to stimulate critical thinking, explore possibilities, and discuss the logical or philosophical implications of this important and timely topic with your students.
Knowledge and the understanding and application of that knowledge is power!
Digital Food For Thought
The White Hatter
References:
1/ https://thewhitehatter.ca/blog/chatgpt-friend-or-foe-what-parents-educators-need-to-know/
2/ https://openai.com/product/dall-e-2
3/ https://thewhitehatter.ca/the-new-use-of-artificial-intelligence-ai-to-commit-crime/
4/ https://www.vice.com/en/article/pkadgm/man-dies-by-suicide-after-talking-with-ai-chatbot-widow-says
6/ https://www.priv.gc.ca/en/opc-news/news-and-announcements/2023/an_230404/
7/ https://cyberbullying.org/generative-ai-as-a-vector-for-harassment-and-harm
10/ https://www.informationweek.com/big-data/generative-ai-and-building-a-more-trustworthy-digital-space