The Excitement and Reality of AI in Education

I believe it is safe to say that we are in the midst of the 5th EdTech Revolution, brimming with innovative ideas for the education space.?

And as the CIO of one of the largest districts in the U.S., I've already seen firsthand some of the promise and pitfalls of AI in education.?

The potential benefits of AI are enormous, and we're closer than ever to embracing my personal philosophy which I call “BOLT” - Building One Less Thing.?

This concept is about creating tools and resources that simplify rather than complicate the lives of students, teachers, staff, and leaders. It's about achieving Maximum Impact with Limited Resources (LRMI), about "one less thing to do" instead of "one more thing to do."

I am confident that AI can help get us there, with estimates showing that AI could automate 20 to 40 percent of current teacher hours spent on non-instructional activities, potentially saving around 13 hours per week.

Not Every Great Idea is a Company or Product

However, not every idea is viable for business or product development. And importantly, not every product labeled with "AI" truly embodies artificial intelligence.

I have sat through numerous “AI” demos over the past couple of years, some great, and some that are nothing more than an empty shell UI or a previously existing product with an API integration to OpenAI .

I have even sat through demos where companies tried to use my interest to gain seed funding from investors.

None of this is inherently good or bad.

But we have to be careful not to let the hype of AI change our decision-making process around solving problems of practice,

And we cannot sacrifice student and staff data, privacy, and security. Period.

Be careful of MVPs

Minimal Viable Products (MVPs) are a necessary part of the evolution of companies and products.

And I cannot think of a significantly used EdTech solution that did not start as an MVP prior to reaching full scale.

However, not every problem/school/district is a match for an MVP.?

MVPs take an ability and willingness on behalf of the school/district to have things go wrong and to help build the product into a more robust solution.?

Schools/Districts must also be willing to take the risk of the MVP failing and the company going out of business.

I will be the first to say that I have brought in some amazing MVPs that have been very successful and gone on to be incredible full-scale solutions.

If there is a problem set that fits and the school/district understands that a product is an MVP and what that means to both organizations, then it can be a great fit.?

The New World of AI MVPs

That said, AI MVPs bring a whole new level to the risk-reward ratio. While they offer potential benefits, they also come with significant risks, particularly in educational settings where the stakes are high.?

The data privacy and security concerns associated with AI in education are profound.

AI-powered tools require vast amounts of data to function effectively,?and this can lead to unintended consequences, such as unintentional data exposure, biased or inaccurate AI models, and legal and regulatory compliance issues.

To mitigate these risks, it’s crucial for schools/districts and educators to adopt strategies such as developing data literacy, implementing data collection guidelines, conducting privacy impact assessments, and promoting a culture of privacy awareness.?

Clear data privacy policies and regular assessments of AI algorithms are essential to ensure the fair treatment of students and compliance with privacy laws.

We must also be careful of the implications of poorly built AI products.?

Feeding AI models with irrelevant or low-quality data can lead to biased or poorly performing models, which may deliver suboptimal learning experiences and widen existing learning chasms between students.

If they have not already done so, each school should adopt appropriate technical and security measures to safeguard personal data and address algorithmic biases, ensuring the potential of AI is harnessed without compromising privacy rights.

Let's Sum It Up

While the excitement around AI in education is warranted, we must proceed with caution.?

AI MVPs in education require careful consideration of their fit with specific problems, schools, and districts.?

They demand an understanding from all stakeholders that these are evolving tools with inherent risks.?

The ultimate goal should be to leverage AI for Limited Resources Maximum Impact (LRMI), but not at the cost of compromising student and staff data, privacy, and security.?

The challenge is to find the right balance between innovation and responsibility, ensuring AI enhances rather than detracts from the educational experience.

#whateverITtakes