The impact of AI on Education (and why Plato would have liked it)
Generative AI has the potential to revolutionize education by making learning more personalized and accessible.
As we learned at school, Plato had reservations about writing. One of his key criticisms was that a written text cannot respond or defend itself when questioned because it is unidirectional and not “interactive.” In Phaedrus, Socrates says:
“I think writing has this strange feature, which makes it truly like painting. The offspring of the painting stand there as if alive, but if you ask them something, they preserve a quite solemn silence. Similarly with written words […] if you ever ask them about any of the things they say out of a desire to learn, they point to just one thing, the same each time […] in the presence both of those who know about the subject and of those who [don’t].” - Penguin Classics edition, translated by Christopher Rowe.
Today, for the first time in history, with Generative AI, a text can “reply”. Imagine learning “directly from Plato” through a chatbot built on his dialogues and some of the best works discussing his philosophy. Of course, there are caveats: for the chatbot to be reliable, authoritative, and free from hallucinations, it must be grounded in quality content. In fields like philosophy, quality content mostly comes from trusted publishers. Therefore, authoritative publishers will remain central to the educational ecosystem for the foreseeable future.
The ability of AI models to create content, reproduce human-like understanding, and adapt to individual needs presents both opportunities and challenges for education. Let’s explore some of the key impacts.
Personalized learning experiences
As Salman Khan, founder of Khan Academy, highlights in his book Brave New Words: How AI Will Revolutionize Education (and Why That’s a Good Thing), published in May 2024:
“Educators have known for millennia that one-on-one tutoring that works with students at their own time and pace is the best way for people to learn. It is what Alexander the Great had with his teacher, Aristotle […and] today, top athletes and musicians continue to learn through one-on-one coaching.”
“In the eighteenth century, we began offering mass public education to everyone. We didn’t have the resources to give every student a personal tutor, so instead we batched them together in groups of thirty or so, and we applied standardized processes to them […]. That system dramatically improved the overall level of education […]. Still, the approach isn’t optimal for the majority of students.”
Generative AI can finally enable personalized learning. AI can analyze a student’s performance and tailor educational content to their specific strengths and weaknesses, much like a human tutor would. AI systems can also generate practice exercises, create multiple versions of tests, and adjust content in real time. Adaptive learning platforms, powered by AI, can track student progress and dynamically modify the curriculum to match their needs. This degree of personalization was unimaginable till a few years ago.
Educational publishing at the crossroads
Educational publishers are at a critical turning point. The digital era had already disrupted the traditional textbook-centric business model. For example, in 2021, when Pearson launched its subscription service Pearson+, then-CEO Andy Bird emphasized the company’s shift toward becoming a leader in lifelong learning, offering engaging digital learning experiences. However, Generative AI is accelerating this transformation. Leading publishers are integrating AI capabilities into their platforms to enable dynamic content creation, personalized learning paths, and automated assessments. While textbooks won’t disappear overnight, they are no longer sufficient.
The convergence of publishers and educational providers
This point is so significant that it deserves a more in-depth article. For now, suffice it to say that educational publishers can no longer focus solely on books—they must develop AI-powered educational tools. Meanwhile, educational providers will increasingly need high-quality proprietary content to power their own tools. This convergence between publishers and providers will only grow.
Better (and cheaper) content creation
Generative AI is transforming school book publishing by automating the creation of textbooks, lesson plans, and supplementary materials. AI processes vast amounts of data to produce up-to-date content that reflects the latest developments in any subject area. This not only reduces publishing time and costs but also ensures that educational materials remain current.
Enhanced accessibility
Generative AI can make education more inclusive. For students with disabilities, AI tools can provide alternative content formats, such as audio descriptions, simplified texts, or visual aids. AI can also translate educational materials into multiple languages, making learning accessible in diverse, multilingual classrooms.
Teacher support and workload reduction
Educators benefit from AI tools that automate administrative tasks such as grading, attendance tracking, and report generation. Generative AI can suggest lesson improvements, create teaching resources, and provide tools to enhance instruction. By reducing routine workloads, teachers can focus more on student engagement and personalized teaching.
Students’ assessments
Generative AI is reshaping how assessments are conducted. AI-powered adaptive testing can provide real-time feedback and grades, and adjust question difficulty based on student responses. However, this also raises challenges around academic integrity, requiring new strategies to prevent AI-assisted cheating.
Preparing students for an AI-driven world
As AI becomes increasingly widespread, education systems must prepare students to thrive in an AI-driven future. This involves teaching digital literacy, critical thinking, and ethical considerations regarding technology use. Integrating AI into education not only enhances learning but also equips students with skills essential for the evolving job market.
Like it or not, there’s no turning back, so it’s better to (carefully) embrace AI in education and use it to its fullest potential.
🛠️ Major AI model releases in 2024, organized by developer:
OpenAI
GPT-4o: A multimodal model capable of processing text, images, and audio. It supports real-time interaction and tasks like sentiment analysis, video analysis, and audio generation.
Sora: A text-to-video model that generates videos directly from prompts, enabling creative applications in video production.
Google
Gemini 2.0: A multimodal agentic model designed for autonomous task execution with advanced reasoning and tool integration. It processes text, images, and audio while generating multimodal outputs.
Gemini 2.0 Flash: A low-latency version optimized for developers with enhanced performance.
Meta
Llama 3.3: A language model optimized for multilingual tasks, coding, and reasoning with a focus on efficiency.
Motivo: A motion-rendering AI for generating animations for avatars, enhancing VR and Metaverse applications.
Anthropic
Claude 3.5 Sonnet: An updated language model with improved performance in coding, workflows, and chart interpretation. It includes the ability to preview code in real time.
Stability AI
Stable LM 2: Open-source models with parameter sizes of 1.6 billion and 12 billion, aimed at improving AI application capabilities across industries.
Mistral AI
Mixtral 8x22B: A sparse Mixture-of-Experts model with 141 billion total parameters but only 39 billion active at a time, improving cost-performance efficiency.
Microsoft
Phi-4: A compact language model excelling in mathematical reasoning and problem-solving tasks.
AI21 Labs
Jamba: A hybrid architecture model with a context window of 256K tokens, supporting extensive input processing.
Inflection AI
Inflection-2.5: A conversational AI model achieving high efficiency by using only 40% of the training FLOPs compared to GPT-4 while maintaining competitive performance.