My niece is exceptionally curious and intelligent. That is one of the key reasons why my sister and her husband pay the cost of private tuition to send her to a school where she learns in a classroom of six students per teacher.
This fortunate privilege grants her mentorship opportunities with her teacher and a positive relationship to education as a whole—a gift to her future and a fortune for some public education classrooms that now only seat 24 students instead of 25.
But why does my niece deserve this experience over another child with the same or greater curiosity and intelligence?
Crowded public schools can be good for some things. Rubbing shoulders with kids from diverse social, ethnic, and economic backgrounds is a great way to gain perspectives and insights into how the world works.
Not that all public schools are the same. Even within the same district, the quality of education, the number of kids per class, safety, and available extracurricular programs can differ widely.
And that is the issue with education today.
Privileged families can ensure that their child receives the best education possible, either by paying for private school or by renting an apartment “across the train tracks” so their child attends the better public school.
None of this speaks to the teachers who likewise struggle, doing their best with the resources available – which are not enough. The education system expects teachers to be educators, babysitters, intervention specialists, and psychiatrists to the hundreds of students they interact with daily.
Test scores are falling and more and more children cannot read, write, or perform simple arithmetic. College degrees are grossly expensive and are typically required only to help filter out job applicants without being necessary even to perform most jobs.
The education system we know today is exclusionary, promotes inequality, can be unsafe, and does not provide outcomes that serve the next generation of citizens, families, voters, or workers.
And that is why it is time to think beyond today’s education system and the new opportunities afforded by technology.
Enter education powered by AI, an innovation that has the potential to provide every student with a better education than the one they have now. One that will not be without problems, but an improvement over what we endure today.
The Role of AI in Education
How will AI revolutionize education?
Education AI (EdAI) is an AI tool powered by large language models (LLMs), not unlike ChatGPT, Claude, or Anthropic. The key difference is that EdAI might be more confined to specific education needs.
Tools like ChatGPT are trained on large amounts of data from recorded human knowledge and experience. The company has not revealed precisely what data it uses to train its models or where that data comes from, but they are now battling several copyright infringement lawsuits.
Access to so much data and complex algorithms can trick you into thinking you are chatting with a human. None of this means an LLM is intelligent or capable of sentience. Tools like ChatGPT parrot information.
We can not ignore the fact that tools can do impressive things, but that does not mean they are well suited for a classroom environment.
For example, AI has been caught teaching people how to acquire ingredients to make a bomb, make poison, or build a script to hack into a web server.
Some curricula might touch on those topics, but most will not.
That is why we want to think of EdAI as something more akin to an encyclopedia—a tome of information carefully developed and curated.
To that end, EdAI is not a replacement for human teachers.
EdAI will provide every teacher with enough teacher’s aids so that each student will have an AI assistant to help them learn. The assigned teacher will oversee each of these AIs. This frees up the teacher to focus more energy on human tasks, contributing feedback to the people who maintain the EdAI tool while also communicating with parents and administration.
Teachers using EdAI might start their morning reading a dashboard with data about students. These students might work entirely from home or spend part or all of their day in the same building as the teacher. Others could travel with their families across Europe, never missing a beat as they learn remotely.
No student is not necessarily working at the same pace either, with one student at the age of six already studying from a medical textbook. While another student of the same age works on identifying shapes. Now imagine a student at the age of 12 shadowing a plumber for ten hours a week.
From data and human input, the EdAI will provide analysis about each student’s progress, including recommendations for strategy or other comments.
From there, the teacher uses their own experience and knowledge to develop strategies that support the specific needs or goals of individual students. This might include mentoring, offering guidance in subjects where the EdAI tool is not working well, or escalating issues to administrations and outside health and service providers to address observed health or safety risks.
Ultimately, the idea is to eliminate all set schedules and curricula. There are no standardized tests because those tests would be a measure of nothing. All students are learning different skills and knowledge at different times, and how each student will apply those skills and knowledge will be unique to them.
Imagine a student who has expressed interest in plumbing. EdAI would create interactive modules that challenge the student to apply mathematical thinking to make cuts for an under-sink pipe. Comparatively, a student on a path toward a medical field may be challenged with converting dosages of a drug based on a patient’s weight.
Some universal testing might still exist. People need to know how to manage finances, fill out a basic form, drive, cook, and clean.
In the same vein, we still want to encourage extracurriculars. Spending all your time indoors with social media and video games harms human health.
Some of the data EdAI collects will relate to information about a student’s participation in sports, volunteer activities, and time spent socializing. All will be both important for their education but also the specific goal of creating more community in society, which has been in decline.
That is the big-picture pitch. We agree that education is broken. AI is the leading technological innovation that, if implemented and used appropriately, can propel humanity into a more desirable future for this generation and those who come after.
But we have to be upfront about the potential issues with AI. These tools will continue to proliferate throughout our lives, the economy, and the government. Good and bad things will come from that.
The following sections of this article dig deeper into what EdAI means for education, what AI will do best, and where it could all go wrong.
Read, think, and message me whether you agree or disagree with what our education system needs to better prepare the next generation to thrive in this world.
What EdAI Will Do Best
Tools are only as good as those who design, build, and use them. EdAI will not emerge fully formed in education. LLMs and the algorithms that power them are still relatively young, requiring large amounts of computing power and energy.
There is a lot of work to be done, but the work that already exists has revealed some early possibilities for what AI might accomplish.
When we imagine how EdAI will serve all aspects of education, we need to be both hopeful about what it might do and realistic about where it could go wrong.
The following sections aim to describe this in as much detail as possible.
Resolving Disruptions and Dynamic Learning
At the surface level, EdAI solves one of the most frequent complaints from teachers: classroom disruptions.
A teacher starts their lesson plan, reading the first sentence of the material, and then a hand shoots up to ask a question. The student is already confused and refuses to wait to hear what else the teacher has to say.
The teacher moves on, but another student blurts out a question of their own. The class is now rowdy, other children begin talking, and those with attention disorders are lost for the next hour.
The teacher moves on to testing. One student lives in a household with parents who do not speak English, and the teacher has to make time to help them with the test while other students struggle in another way.
To compound challenge upon challenge, a student with a diagnosed or undiagnosed condition starts running around the room, ripping posters off the way. Others have no home. Some are hungry. There may be abuse. Emotional issues lead to behavioral dysregulation. And so on and on and on.
Meanwhile, other students sit quietly, their attention focused and ready to learn.
The root cause of these issues is that students of the same grade are not at the same developmental stage, they have different needs, and there are simply too many of them sharing the same space.
In schools today, the curriculum is not being taught, and students are unprepared for year-end testing. Poor test scores reflect poorly on the teacher, the school, and the student. Parents are mad at the school district, the school district criticizes the principal, and the principal retargets the teachers. If we are lucky, the teacher does not pack up and leave right then.
This leads to a poor experience for all involved.
How does EdAI fix this? Return to the earlier situation of a student raising their hand to ask a question after hearing the introductory sentence. Rather than disrupt the entire class, the student learns to ask those types of questions in their EdAI tool first.
If this fails to resolve the issue, the teacher might be alerted for approval of an alternative lesson plan. The EdAI then begins a lesson on how topic sentences lead into paragraphs and larger ideas. Had the student listened a little while longer, they might have had their question answered.
The deviation from typical coursework seeks to immediately address a student’s need without involving anyone else. Furthermore, this deviation might inform the EdAI to add a skills challenge to evaluate how effective this deviation was for the students and to reinforce what they have learned.
Because this is all dynamic, a teacher or EdAI can add complementary lessons, such as lessons on note-taking, memorization challenges, or going back to the basics of how we form and communicate ideas.
The students who were described as having sat quietly and ready to move on to the next lesson could shift to an accelerated path. They remain on that path, only deviating if they begin to struggle. Again, at that point, the lesson plans slow down, or supplemental learning is added.
The goal is to create higher levels of satisfaction for everyone. Teachers can focus more on teaching and seeing children succeed. Students are being given that hands-on education they may or may not need, learning at a pace that suits them and, ideally, boosting learning outcomes.
We also try to limit or eliminate interruptions that can impact every other student in the classroom. As a bonus, we set a goal to boost teacher retention and interest in the field.
Real-Time Analysis of Student Health and Safety
The power of EdAI lies not only in its expansive access to information and its flexibility to adjust schedules and coursework but also in its ability to assess students based on what it sees and hears.
As we saw with ChatGPT-4o when first unveiled, and presumably with future iterations of this model and models from competitors, AIs can be trained to understand the real world, from speech to colors, shapes, animals, and even our posture.
Apply this real-time analysis to the classroom to provide deep insights into student needs beyond raising their hand or how well they perform on a test. What can be learned from studying physical movements, facial expressions, or changes in voice?
A student might not ask questions because of shyness. What is the underlying cause of shyness, and what lesson might encourage them to be more expressive? The tool identifies a hint of a speech impediment that is not obvious to a teacher’s ears. A student may be getting tired, their eyes dropping slightly, which an AI could detect, and they are asked why they did not get enough sleep (do they have a place to sleep or did they sit up all night playing video games).
This makes an AI superior because it can focus 100% of the time on watching and listening and does not carry any emotional burden due to constant exposure, unlike a teacher who knows that asking difficult questions and raising the alarm can be taxing and lead to a lot of difficult inquiries.
Note Taking, Summaries, and Practice Tests
AI tools are already being used in classrooms to help students answer test questions and write their essays. Cheating is an issue that will require changing how students take tests and issuing clearer guidance on what happens if they are caught cheating.
But when we look past the issues with AI taking the work out of education, there is value in allowing a student to use a recording app on their phone and allowing an AI to transcribe the lecture, rewrite and summarize information, and generate test and essay questions to practice with.
Unfortunately, there are many gray areas in academia. While one teacher may approve of using AI to summarize information, others could argue that listening and identifying important information, disseminating that information, and creating flashcards or essay questions manually are too important of a skill.
Having different expectations and rules will introduce confusion. Whatever the decision around these tools, it should be universal and clearly stated that punishments will be imposed for those who use the tools to circumvent actual work and learning.
Another advantage to consider is how a teacher can pre-record their lecture and distribute it among students so they can engage with the material using the EdAI to add sources, images or videos, and other supplemental information. At the same time, teachers could review how students engage with the material, informing the teacher of how they might want to modify the lecture to clarify confusion or make it more engaging or educational.
If the EdAI is proficient with language translation, the potential for a teacher to provide their course to students in different parts of the world could unlock new opportunities for those in areas that would otherwise not have the chance.
Guiding Friendships and Interests
The introduction of this article discussed the notion that a classroom is a good place to develop social skills quickly. Having peers of diverse backgrounds creates adversity and raises self-awareness. A peer can quickly check if you say something they perceive as offensive or if they simply disagree with a particular viewpoint.
Inversely, my niece attending a small private school will likely see her situated among students of relatively equal status.
At its core, neither idea is right or wrong.
Some parents may never want their children to face “the realities of the world” beyond the safety net they work hard to create around the family. These realities can relate to socio-economics, religion, or politics.
Being around only similar people can also be problematic. The internet is one example of this. Before Discord and Twitter, the people you socialized with the most were probably your neighbors and coworkers. You had no choice in who these people were, their opinions, and political or religious leanings. And you found they were friendly people who you got along with well.
With the internet, we have seen the rise of people isolating themselves within like-minded groups give way to mob mentalities that grant them the ability to fight effectively against their “enemies.”
The rise of partisanship and ideology at all levels of society is leading to a rise in conspiracy theories, mistrust of science, and the belief that the government is an active enemy.
EdAI can potentially isolate or unite students. So do we intentionally “mismatch” kids so they can see how the other side lives? Are there inherent risks of bullying or violating a parent’s personal beliefs?
It would be easy to say that AI should be used to identify and categorize students based on a range of factors and then create circumstances in which any two students are most likely to interact. But that seems like a great way to introduce bias into our system.
Yes, ensuring students are exposed to opposing views and situations is important for gaining perspective on life. Adversity is said to build character. But it could go too far too easily.
Ultimately, we want to unite people through conversation and shared interests while exposing students to various experiences. If we are going to support democracy and encourage the public to engage with their government, we should want to encourage students to go out and learn about themselves and others, how to share their beliefs, and have healthy and productive conversations.
EdAI might help to facilitate that, but it may be an area we label as out of the scope of what it should be doing.
Scalable Regardless of Age or Life Situation
EdAI should scale to any age or level of education. For older students, it would match their mannerisms to create an environment that does not feel intimidating or condescending.
This will help those who have already completed their education, dropped out before finishing their GED, or never went to college.
Providing a tool that scales in this way is essential. In part because K-12 will not define the future of education.
In our careers, with future advancements in the digital space coming faster and faster each year, new and varied skills will be necessary to do most jobs. What you know today may not be relevant in just a few years.
The younger generation that grew up with EdAI may have an easier time adapting to the ever-changing landscape of their careers, jumping between companies or jobs with the assistance of a virtual assistant that may be supplemented with a real teacher.
Much like the generation of adults who began their careers on the typewriter, they came in one day, and a computer was on their desk. They had to learn how to use an operating system, connect to a network, and use formulas in a spreadsheet.
The difference is that the transition to computers in the workplace was relatively slow. The rate of transition to AI in the workplace will be rapid. So, we risk leaving behind older individuals who already face ageism.
This makes it important to consider how we can expand access to all age groups and provide free training resources to anyone who wants to use EdAI.
The Risks and Challenges of Education AI
There will be blind spots and opportunities for malice within the widespread adoption of EdAI.
EdAI works because of the vast libraries of data it contains. Over time, adding or removing data will become necessary. Much like textbooks are updated over time, so too will EdAI.
But what data do the people in charge add or omit from EdAI? Do they omit things from history because they contradict a held political ideology? Do we avoid things that are simply too disputed?
The danger of AI versus a textbook is that changes to a textbook are obvious. The complexities of an LLM mean that any changes may be overlooked entirely.
AI tools are also ideal for spreading misinformation and generating deep fakes of people. There are also fears that they will diminish higher-level thinking as people use AI to answer emails, write outlines, or make business proposals.
And those are only a few of the risks. AIs can provide inaccurate, outdated, or flawed responses, even when the data they contain is accurate. There is also a risk of privacy violations if the information you input into the AI is used to train future versions of the AI. This includes banking information, social security numbers, or details about when you will or will not be at home.
Before a rollout of EdAI can happen, policy-makers and education institutions need to elect non-partisan representatives to discuss and advocate for solutions to the issues outlined in the following sections.
Bias in AI and Winston Smith
Controversy has long surrounded education, with decades-long fights over race and religion, the basic foundations of science, who won a political election, and the effectiveness of vaccines.
This begs the question of who do we put in control of EdAI?
Does the state or federal government have a say in what data is used to train the EdAI or how it is tuned to respond to students?
Will a political candidate run on the promise of installing a pastor of a Christian denomination to have a say in how the algorithm returns information to teachers or students to ensure a pro-Christian message?
Much like with book bans, can a group of concerned parents rally to get entire portions of an AI’s functions banned?
How easily can someone log into the tool and make changes?
Orson Welles’ book “1984” warned that when you give someone control, they will seek to use that power to change established truths and history to serve their aims. For many years, people have sounded the alarm bells that 1984 has come true. Parallels crop up from time to time, but none ever seemed like a great worry.
But in the era of AI, it is becoming easier for bad actors to use AI to spread misinformation via social media and legitimate news channels. This is made worse now that AI can generate convincing images, videos, and audio in seconds.
The ability to manipulate students, teachers, and parents with information that seeks to empower certain political groups or sow discord is easy to imagine.
For example, a school district limits access to queries such as, “Does systemic racism still exist?” They do not prevent the query, but they do apply a filter, ensuring the AI responds with a decisive “No” while encouraging students to reflect solely on how “Confederate states empowered African immigrants with land ownership, economic opportunities, and a civilized standard of living.”
It is equally easy to imagine a school district teaching the King James Bible version of creationism using generated videos of Adam and Eve walking out onto the screen and having a conversation with the students. At one point, they tell the students that this is a live feed from heaven, and God wants everyone to go out as warriors to spread the truth and be free of democratic tyranny. This is before they cut to Darwin standing in front of a crowd of academics, tomatoes flying at his head, while he admits to being a fraud.
While these are potentially extreme examples, we can not underestimate the will of people to spread anything that goes against academically sound viewpoints.
And there may not be anything that can be done to prevent this. The courts have shown to be unprepared to take on the burden of these cases, such as teaching creationism or banning evolution, which can go on for decades without resolution.
Education That Violates Privacy and Data Laws
Stealing data, invading privacy, and selling personal information are issues in digital spaces, but they will only worsen with AI.
It will be worse for children raised alongside AI, much like Gen Z, who have never known a world where you do not take and share personal updates or photos and videos via social media daily.
The next generation will be born with AI at their side. They will feel comfortable with virtual assistance being their teacher and imaginary friend.
All to build trust that could be easy to exploit.
Much of this will become possible thanks to adding more memory within the AI. Currently, AI has limited memory, meaning that a piece of information shared with the AI a few days ago might not be referenced by the AI today. However, increasing that memory means storing everything about a child, from birth to death. In the end, the AI knows more about an individual than they even knew about themselves.
If someone gained access to this data, it would be a treasure trove for marketing companies and scammers.
Despite efforts by governments to limit how data is collected and used, people can breach firewalls to mine data from servers and then use that information to violate advertising and privacy laws. Hackers can worm their way into computer webcams and microphones. A predator can learn a child’s bus route and figure out a parent’s work schedule.
We have even seen large, public facing companies using data to exploit children blatantly. Epic Games, the maker of Fortnite, was sued by the Federal Trade Commission for gathering data on children and intentionally creating systems that “violated the Children’s Online Privacy Protection Act (COPPA) and deployed design tricks, known as dark patterns, to dupe millions of players into making unintentional purchases.”
There have to be guardrails. But we should also consider that a middle ground can exist. One where parents agree to sell their data with identifying information removed.
For the sale of this data, the school system receives funds, and the family receives legitimate discounts on offers such as airfare, cruises, a college education, specific offers for housing, retirement, investments, and the like. Think of it as a new form of fundraising and currency. Instead of candy bars, it is your child’s favorite color and television show.
Perhaps in the future, how we think about data will differ, and our concerns will disappear. But until that time, violations of privacy and data collection could steamroll any attempts to get EdAI off the ground, so it is an issue that will need to be addressed sooner rather than later.
Digital Poverty and Exploitation of Poor Countries
We take for granted that in developed nations, especially in America, some corporations can offer six and seven-figure compensation deals to attract the best talent from around the globe.
This endless coffer of money does a few things.
First, it deprives less developed nations of those individuals who, had they not been lured away, might have built their ideas in their home country. This would support the local economy, help to develop new talent, and contribute to that ‘s economic standing.
The second is that companies are less likely to invest resources to expand operations or training to areas of poorer countries voluntarily. America and China are pumping vast sums of money into AI hardware and software and the companies and governments within these countries are vying for limited resources currently.
Due to its cost and domestic demands for this technology, EdAI may wind up with limited global access. Part of this is a lack of infrastructure to support energy production or server capacity to run the tools in a way that makes financial sense. The other is that companies do not want their intellectual property taken by government mandate.
This can lead to digital poverty. In education, a situation arises where the tools and technology to enable EdAI are technically available to everyone, but many of the factors mentioned in this section restrict its availability, effectiveness or usability.
Given the rates of progress in AI today, the expansion of AI across America and China may mean that children in less developed nations fall even further behind in the world than they already are and have little hope of catching up anytime soon.
In the same vein, perhaps these tools will be rolled out to those countries. But it could be without good intent. Such as using the tool to gather data and violate privacy, to sow political discord and destabilize the local government, or to more efficiently identify and exploit vulnerable children.
Incorrect Information
AI tools are only as good as their data, their algorithm, and the person querying the tool. Moreover, LLMs are fundamentally predictive models. When you make a query, the AI tool tries to understand the context of the query, and then, word by word, builds sentences, paragraphs, and entire ideas.
Going word to word relies on “weights” that help the AI to choose the best possible word. These weights are programmed into the AI by engineers and people working as “prompters” who rate responses.
Due to how LLMs work, there is always some probability that a response can be wrong or inappropriate. Everyone gets things wrong now and then. The problem is that when a teacher gets something wrong, it is easy to raise the concern and have the teacher address it. Most people, when incorrect, will acknowledge their mistake and seek to change.
An AI can tell you it was wrong but will continue to make the same mistake. It has no empathy and can not suffer from being fired or jailed.
The spread of AI in society is a contentious issue. For that reason, there will be heightened scrutiny over whatever EdAI is at launch and possibly far into its lifespan. So much so that there could be widespread accusations of malice on the part of the system’s designers, leading to conspiracy theories, banning the tool entirely, or allowing people to gain control over the tool who should not be allowed to be in control.
There will be opportunities to overcome this by informing people early and often of how EdAI works and ensuring there is always a feedback loop for people to submit concerns. If you have used ChatGPT, you will see that a user can rate the output of an AI.
Unfortunately, because adding or removing information from an AI requires retraining the entire model, it is not guaranteed to be fixed. So, there needs to be an interim solution that allows the AI or teachers to intervene when there are any concerns over the accuracy of a response.
Violations of Copyright and Ethics
AI companies are violating the copyrights of publishers, writers, and artists, but the courts may rule that they are not.
There have been interesting discussions around this. After all, humans study, emulate, and cite writers, scientists, and artists in their work constantly. In that way, it is transformative and, therefore, Fair Use. This is the argument AI companies are using currently.
Copyright law was not written with consideration of how AI is trained or works. These issues need to be resolved before EdAI can roll out. No institution wants to allow students and teachers to use a tool to create essays, stories, or research papers that might attract the attention of a copyright holder who then sues one of the involved parties.
Loss of Foundational Skills and Agency
Human agency is reduced anytime another person, a machine, or a company supplies some or all of a necessity. Whether that is the grocery store replacing farming your own land for food or hiring a ghostwriter to write your next novel.
Human thought and creativity are the areas most at risk of being lost to AI. Generative AI tools allow us to summarize content, create outlines, develop arguments, and design finalized artworks without the need to learn anything.
While AI tools can automate repetitive tasks or accelerate workflows, many argue that specific tasks, such as reading a book and writing an essay defending a point made by the author, are repetitive and not necessary.
But it is necessary. Ask ChatGPT, “Who was the greatest Roman emperor?” Most of the time, the response will state that Marcus Aurelius was the greatest Roman emperor. There is a chance ChatGPT will deviate, but the probability is low.
ChatGPT responds with this answer because in most popular literature and articles, the authors all state that Marcus Aurelius was the greatest Roman emperor.
Many people agreeing with someone does not make the response right or good. All this does is reinforce established and popular viewpoints, limit the further pursuit of diversity in research, and reduce creativity and thought across all areas of society.
For that reason, AI can be a boon for education. Yet, it can diminish education by making people less involved in learning and challenging existing viewpoints.
Lose of Human Connection
A threat of AI is the unknown that comes from removing human connection. What happens to a student who speaks with an AI all day?
Could this increase rates of depression, diminish procreation, shift expectations for dating as a result of AI girlfriends becoming more common, give rise to emotional issues, or develop yet unknown disorders?
While we may no longer need large school buildings to accommodate every student, as many may be attending class from home, we still need to encourage students to go out into the world and enter communal spaces, take part in local athletics, the arts, and find opportunities to contribute to the local community.
Highly Educated Populations With Zero Opportunities
Developed societies have deemed that education is vital to a country’s success. For several decades, this has meant steering students away from blue-collar work into office jobs and STEM fields.
The result is generations of over-educated and underemployed individuals with college debt, a high dissatisfaction in their careers, and millions who have simply given up on being in the workforce.
Education has helped many find new opportunities, network, and create careers with high earning potential. But clearly, not everyone can succeed in the same way. We need ditch diggers, gas station attendants, and delivery drivers.
Is it possible to create a new system of education that seeks to be wholly optimized for each individual? Society does not have enough work to give everyone the opportunities they want.
That is not to say new opportunities will not emerge, but we may have to accept that there will be greater competition for many roles, which can lead to feelings of dissatisfaction or resentment.
Thinking Decades Ahead
There are many practical uses for AI in society.
A great example is a company that goes to dermatologists and buys all of the photographs they have taken of moles, skin tags, melanomas, and the like. They then feed that into an LLM to create a tool that makes detecting the risk of cancers or other ailments more accurate and possible from your smartphone at home.
Thousands of such innovations will emerge over the next few years. Their positive impacts will be wide-ranging, from accelerating the economy to bolstering existing jobs.
When EdAI arrives is anyone’s guess. The first versions are likely to come from a privatized, for-profit corporation. The government could offer a stipend to those who want to move their children into that system, allowing the government to pivot out of education slowly while AI companies figure out how to make these tools work.
Whether that ends well or not is up for debate. The argument goes that corporations do things more efficiently than governments, but then we just have to look at higher education and see that the costs and benefits of getting a degree no longer work.
At the same time, the government hardly understands new technology. Congress is only now talking about banning TikTok in 2024 when there should have been regulations and restrictions on all social media platforms years ago.
With so much at stake, a lot of work needs to be done in the coming years to implement a plan that allows EdAI to succeed and serve primarily as a benefit to humanity, not those who would seek to use it for individual benefit.
What Students Need for EdAI to Succeed
Students need an AI tool that helps them achieve their potential. However, students must not feel specific expectations for their future. We need blue—and white-collar workers who perform a variety of tasks. We also need individuals who are active in the political process and who help to make communities safe and happy.
For that reason, this tool needs to serve them outside of education. From the early stages of life, children will be introduced to AI like they would a teddy bear.
It is likely that the children who will be born over the next few years will grow up intertwined with AI that follows them, learns about them, and almost becomes a companion – like an imaginary friend with whom the world can interact.
This could have hugely positive outcomes, or it could lead to a dark future of individuals who isolate themselves with their AI, relying entirely on AI to guide them, to reinforce their preexisting beliefs, and never daring to contribute to the economy, the sciences, or the arts.
We need to tread carefully and not rush too fast into this area. Healthy boundaries may be required to understand the implications of AI in any of our lives, and then we may slowly allow it a longer leash.
What Teachers Need for EdAI to Succeed
Teachers need to feel confident that EdAI supports their efforts and does not need to be micromanaged to work. EdAI’s data must also be helpful to provide high confidence that any strategy developed using this data is the right decision.
They also need to know the limitations of EdAI, with clear guidance from and support from administration and approval from parents. They must also know how to communicate concerns about the accuracy of the EdAI and report issues with its functioning.
Most importantly, teachers need to know that EdAI is not a replacement for their work. EdAI does not work if we drop a computer in a child’s lap and trust that the AI will do anything well enough on its own.
What Parents Need for EdAI to Succeed
Parents need to know that regulations for EdAI exist, the tools are safe, the extent to which data tracking is occurring, and that sensitive data is being destroyed.
They also need to feel like EdAI is making them more involved with their child’s education. If we move beyond a traditional K-12 model, parents will need more information to understand if their child is behind, on track, or ahead. EdAI demands flexibility, so parents need to understand their options and that the right decisions are being made.
What Governments Need for EdAI to Succeed
Governments must clarify several things, including reinforcing copyright law with proposals for dealing with AI-generated content. They must establish clear limits for data collection and impose harsh penalties for those seeking to exploit data collection and the spread of misinformation. The government also needs to ensure the promotion of data equality and limit the influence of politicians or private entities on EdAI.
More importantly, we need to overcome partisanship in society. While protecting the rights of people to think and believe what they want is important, governments must stand behind the right to a good education.
The Future of Education
The future of education is best described as adaptive, providing all involved with exactly what they need when they need it.
Students will learn and grow at different rates. They will change careers, hobbies, and interests more frequently, try new things, and have the helping hand of a personal assistant over their shoulder who knows almost everything about them and the world.
Teachers will be teachers. They will find joy in the classroom experience. This will lead to higher retention rates and attract more candidates.
The government and the economy will have access to a workforce that can adapt quickly and does not carry the burden of student loan debt. As all people have access to a high-quality education, equality will rise among the population. Yet, this does not mean everyone will get what they want.
Jobs will become more competitive, with people potentially fighting for a lifetime to get their foot into the door of a career they want.
With EdAI acting as an intervener, society could identify those most at risk of falling behind due to poverty, violence, or drugs and help guide them to a better path.
Those countries with well-functioning EdAI are likely to become even more dominant players in technology and innovation, raising GDP and the standard of living.
Poorer countries will fall further behind.
If AI tools become a crutch in our lives, we might lose higher-order thinking. Society still needs new ideas and solutions, and that will only come if we allow AI to accelerate our existing creativity, thinking, and problem-solving.
Like anything else, AI tools are not the end solution but are a step in the right direction. Indeed, something will come after the AI we are talking about today.
But without EdAI, the traditional education system will not survive on its current path. A lack of funding, poor learning outcomes, and a loss of faith in the college degree mean it is time for something new to replace what we have today.
EdAI can revolutionize education. It will take time because we will not know the results of the first students put through the program for at least a decade or longer.
Given enough time, the issues outlined here will work themselves out. The benefits will become more clear and society sets itself on a path towards the next frontier of human potential.
