Back

Why are silicon wafers round when chips are square?

The surprising science behind the shape of silicon wafers

5 August 2026

Ever looked at a silicon wafer and thought, “Hang on — why is that round?” It’s a fair question, especially when you realise that most of the electronics they power — from your smartphone to your laptop — are full of neat rectangles and squares. So, what’s with the circular dinner plate look? The answer lies in the mesmerising manufacturing process behind one of the most essential materials in modern computing.

From molten silicon to perfect cylinders

To understand why wafers are round, we have to go back to the beginning — literally. Making a silicon wafer starts with growing a single, pure crystal of silicon. This is done using a technique called the Czochralski process (yes, it sounds like a ski jump move). A small seed crystal is dipped into a vat of molten silicon, then slowly pulled upwards while being rotated. The result? A long, cylindrical ingot of solid silicon.

Why circles make the cut

Once you’ve got a perfect silicon cylinder, what’s the logical next step? Slice it up — just like a loaf of bread. These thin circular slices become silicon wafers. Trying to cut out squares would not only waste a lot of material but would also make the whole process incredibly inefficient. So, round ingot equals round wafer — it’s just geometry doing its thing.

The irony of chip design

Here’s the twist: after all that work to create flawless round wafers, manufacturers immediately start carving them up into tiny rectangular chips. It’s the world’s most expensive geometry lesson, and yet, it’s the most practical way to mass-produce the silicon components that power our digital lives.

Want to know more? The Craig’n’Dave Lesson Hacker, explains all.

For more Lesson Hacker videos, check out the Craig’n’Dave YouTube playlist HERE.

Visit our website to explore more cutting-edge tech news in the computer science world!

Related posts

Beyond the IDE: programming lessons in the age of AI

One of the most persistent challenges in a programming classroom is knowing what every student is doing at any given […]

23 October 2026

If a NOT gate turns zero into one, where does that one come from?

How does a NOT gate turn a 0 into a 1 without creating something from nothing? Discover how voltage, transistors and digital logic work together to power the technology inside every modern computer.

15 October 2026

Ofsted after the Big Listen: have the rules changed, or just the packaging?

Ofsted has changed its inspection framework, but many school leaders question whether it has changed the culture surrounding inspection. We explore whether report cards can rebuild trust, reduce pressure and create an accountability system that genuinely helps schools improve.

9 October 2026

Should I make my own resources?

Should teachers really be making all their own resources? We explore the balance between creating, buying and adapting resources — and why thoughtful planning matters more than who made them.

5 October 2026

How does age verification technology work?

Can technology really tell how old you are? Explore the computer science behind age verification and the privacy and accuracy questions it raises.

28 September 2026

UTF-8 is witchcraft, right?

Ever wondered how your computer can happily store a simple English letter, an emoji, a pound sign and characters from […]

22 September 2026

Why teaching models keep returning

Teaching frameworks can provide valuable guidance, but when they become rigid formulas, they can undermine professional judgement. The goal isn’t identical lessons – it’s consistently effective teaching.

18 September 2026

Do GCSEs doom students to fail?

GCSE results may look stubbornly flat, but the reality behind the numbers tells a very different story. Why has further improvement become so difficult, and are we measuring success in the right way?

11 September 2026

Will Quantum Computers Break Encryption?

Could quantum computers really break encryption and put our data at risk? We explore the reality behind the hype, from qubits and Shor’s Algorithm to the future of post-quantum security.

7 September 2026

Back

What Is the Best Material for a Smartphone?

3 August 2026

Why your smartphone is a compromise (not a masterpiece material)
When you ask what the best material for a smartphone is, the honest answer is… There isn’t one. Modern smartphones are not built from a single “perfect” material, but from a carefully engineered compromise balancing durability, performance, cost, and aesthetics.
This is where materials science meets real-world physics — and a bit of marketing.

Plastic: The original tough guy

Early smartphones relied heavily on plastics like polycarbonate. Why? Because it’s light, cheap, and incredibly impact-resistant. In other words, it survives drops that would turn other materials into expensive regret.
The downside? It scratches easily, feels less premium, and doesn’t handle heat particularly well — not ideal for a device that constantly runs like a miniature computer in your pocket.

Metal: Premium feel, technical problems

Aluminium frames solved the “cheap plastic toy” problem almost overnight. Strong, rigid, and excellent at dissipating heat, metal-made phones feel instantly more premium.
But there’s a catch: metal interferes with radio signals. Wrap a phone in metal and you essentially build a mini Faraday cage. That’s why modern designs include those subtle antenna lines — a functional necessity that somehow became a design feature.

Glass: The luxury illusion

Glass arrived to bring style and functionality. Chemically strengthened glass (like Gorilla Glass) resists scratches and enables wireless charging.
But its weakness is dramatic. When glass fails, it doesn’t just crack — it commits fully to the spiderweb aesthetic. One drop and your “premium device” becomes modern abstract art.

Ceramic: The overachiever we rarely see

Ceramic sounds perfect on paper: hard, scratch-resistant, heat-stable, and transparent to radio waves.
So why isn’t every phone made of it? Manufacturing. It’s expensive, difficult to produce at scale, and prone to breaking during production — which tends to upset both engineers and accountants.

So what is your phone made of?

It’s a carefully designed sandwich. Metal for strength, glass for connectivity and charging, and plastic quietly doing the unseen work inside.
Your smartphone isn’t built from the best material. It’s built from the least-bad combination we’ve engineered so far. And then, of course, we all drop it on the kitchen floor to test that theory.

Watch the full Lesson Hacker video on our YouTube channel for the full breakdown.

For more Lesson Hacker videos, check out the Craig’n’Dave YouTube playlist HERE.

Visit our website to explore more cutting-edge tech news in the computer science world!

Related posts

Beyond the IDE: programming lessons in the age of AI

One of the most persistent challenges in a programming classroom is knowing what every student is doing at any given […]

23 October 2026

If a NOT gate turns zero into one, where does that one come from?

How does a NOT gate turn a 0 into a 1 without creating something from nothing? Discover how voltage, transistors and digital logic work together to power the technology inside every modern computer.

15 October 2026

Ofsted after the Big Listen: have the rules changed, or just the packaging?

Ofsted has changed its inspection framework, but many school leaders question whether it has changed the culture surrounding inspection. We explore whether report cards can rebuild trust, reduce pressure and create an accountability system that genuinely helps schools improve.

9 October 2026

Should I make my own resources?

Should teachers really be making all their own resources? We explore the balance between creating, buying and adapting resources — and why thoughtful planning matters more than who made them.

5 October 2026

How does age verification technology work?

Can technology really tell how old you are? Explore the computer science behind age verification and the privacy and accuracy questions it raises.

28 September 2026

UTF-8 is witchcraft, right?

Ever wondered how your computer can happily store a simple English letter, an emoji, a pound sign and characters from […]

22 September 2026

Why teaching models keep returning

Teaching frameworks can provide valuable guidance, but when they become rigid formulas, they can undermine professional judgement. The goal isn’t identical lessons – it’s consistently effective teaching.

18 September 2026

Do GCSEs doom students to fail?

GCSE results may look stubbornly flat, but the reality behind the numbers tells a very different story. Why has further improvement become so difficult, and are we measuring success in the right way?

11 September 2026

Will Quantum Computers Break Encryption?

Could quantum computers really break encryption and put our data at risk? We explore the reality behind the hype, from qubits and Shor’s Algorithm to the future of post-quantum security.

7 September 2026

Back

AI torrenting may not be so legal after all

31 July 2026

AI training is facing a legal reality check as courts begin drawing a clearer line between fair use and copyright infringement. The recent $1.5 billion Anthropic settlement has become one of the biggest warning signs yet for the AI industry, highlighting why data provenance, licensing, and transparency are now critical parts of modern AI development.

The legal reality check for AI training data

Artificial intelligence has developed at an incredible pace over the last few years, with companies racing to build larger and more capable models. But behind the rapid innovation sits a growing legal and ethical problem: where exactly did the training data come from?
The Anthropic case centres around allegations that copyrighted books from “shadow libraries” such as Books3 and LibGen were used during the training of large language models. These libraries contain huge collections of pirated books and copyrighted material, making them highly controversial sources of data.
For years, many AI companies operated in a space where data scraping and large-scale collection practices were rarely challenged in court. That situation is now beginning to change. Regulators, publishers, and copyright holders are increasingly questioning whether AI companies can freely use protected content without permission or payment.
The result is a major legal turning point for the entire AI industry.

Fair use vs infringement: Where’s the line?

One of the most important outcomes from recent court rulings in 2025 is the distinction between transformative use and outright copyright infringement.
In simple terms, courts are beginning to recognise that training an AI model on lawfully obtained material may qualify as fair use because the system is transforming information into statistical patterns rather than reproducing the original work directly.
However, the same protection does not apply when the source material itself has been illegally obtained.

A useful comparison is this:
Reading and learning from a book you purchased is generally acceptable. Building a massive archive of pirated PDFs to train commercial AI systems is something very different.

That distinction could shape the future of AI development for years to come.

The hidden problem: Orphaned data in AI models

One of the most difficult technical and legal challenges surrounding AI training data is the issue of orphaned data.
Once information is absorbed into a large language model, it becomes deeply embedded within the model’s weights and parameters. Unlike deleting a file from a computer, removing specific training data from an AI system is extremely difficult and, in many cases, practically impossible without retraining the model entirely.
A good way to think about it is like making a smoothie. Once every ingredient has been blended together, you cannot easily remove one individual component afterwards.
This creates long-term risks not only for AI developers but also for businesses using third-party AI tools. Even companies that never directly handled copyrighted data could still face complications if the underlying models were trained using questionable sources.
As a result, data sourcing is increasingly being treated like a supply-chain issue. Organisations now need confidence not only in what AI systems can do, but also in where their training data originated.

What this means for the future of AI

The AI industry is rapidly moving towards a more regulated and compliance-focused future.

We are already seeing growing demand for:

  • Licensed datasets instead of scraped or unverified content
  • Stronger data provenance tracking
  • Transparent audit trails for training data
  • Compliance-first AI governance frameworks
  • Greater accountability from AI developers and providers

The “Wild West” phase of AI development is beginning to close. As regulation catches up with innovation, companies that prioritise transparency and lawful data practices are likely to be in a far stronger position moving forward.
The Anthropic case is a strong reminder that AI development is no longer operating in a legal grey area. Courts are beginning to separate legitimate innovation from careless data practices, and the message is becoming increasingly clear: how AI companies collect and manage training data matters just as much as the technology itself.

For developers, businesses, and educators, this marks a major shift in the future of artificial intelligence. Using legally obtained, traceable data may still support innovation under fair use, but relying on pirated or unlicensed content brings serious legal and ethical consequences. The challenge is made even more complex by the nature of AI models themselves — once data has been absorbed into a model, removing it is far from straightforward.

As AI continues to evolve, the industry is moving towards a more accountable future built around licensed datasets, transparent data sourcing, and stronger governance. The fast-moving “scrape first, ask later” approach is rapidly disappearing, replaced by a growing expectation that AI systems must be explainable, auditable, and legally compliant from the ground up.
The biggest lesson? AI is no longer just a technology challenge — it is now a copyright, governance, and trust challenge too.

Watch the full Lesson Hacker video on our YouTube channel for the full breakdown.

For more Lesson Hacker videos, check out the Craig’n’Dave YouTube playlist HERE.

Visit our website to explore more cutting-edge tech news in the computer science world!

 

Related posts

Beyond the IDE: programming lessons in the age of AI

One of the most persistent challenges in a programming classroom is knowing what every student is doing at any given […]

23 October 2026

If a NOT gate turns zero into one, where does that one come from?

How does a NOT gate turn a 0 into a 1 without creating something from nothing? Discover how voltage, transistors and digital logic work together to power the technology inside every modern computer.

15 October 2026

Ofsted after the Big Listen: have the rules changed, or just the packaging?

Ofsted has changed its inspection framework, but many school leaders question whether it has changed the culture surrounding inspection. We explore whether report cards can rebuild trust, reduce pressure and create an accountability system that genuinely helps schools improve.

9 October 2026

Should I make my own resources?

Should teachers really be making all their own resources? We explore the balance between creating, buying and adapting resources — and why thoughtful planning matters more than who made them.

5 October 2026

How does age verification technology work?

Can technology really tell how old you are? Explore the computer science behind age verification and the privacy and accuracy questions it raises.

28 September 2026

UTF-8 is witchcraft, right?

Ever wondered how your computer can happily store a simple English letter, an emoji, a pound sign and characters from […]

22 September 2026

Why teaching models keep returning

Teaching frameworks can provide valuable guidance, but when they become rigid formulas, they can undermine professional judgement. The goal isn’t identical lessons – it’s consistently effective teaching.

18 September 2026

Do GCSEs doom students to fail?

GCSE results may look stubbornly flat, but the reality behind the numbers tells a very different story. Why has further improvement become so difficult, and are we measuring success in the right way?

11 September 2026

Will Quantum Computers Break Encryption?

Could quantum computers really break encryption and put our data at risk? We explore the reality behind the hype, from qubits and Shor’s Algorithm to the future of post-quantum security.

7 September 2026

Back

Why are LEDs so efficient?

The science behind modern lighting

29 July 2026

From space heaters to smart light sources

LEDs are everywhere today — but have you ever wondered why they’re so efficient compared to old-fashioned light bulbs? The answer lies in a clever shift in physics that completely changed how we generate light.

Incandescent Bulbs: Bright, but wasteful

Traditional incandescent bulbs work by forcing electricity through a thin wire until it becomes so hot that it glows. That glow gives you light… but there’s a catch. Most of the energy is actually lost as heat.
In fact, around 90% of the energy in an incandescent bulb becomes heat rather than light. That’s why old desk lamps could warm a room, and why switching them off genuinely helped cool things down.

LEDs: Light without the heat problem

LEDs (Light Emitting Diodes) take a much smarter approach. Instead of heating metal, they use a semiconductor junction where electrons move between energy levels. When they drop to a lower energy state, they release energy as photons — in other words, light.

No extreme heat. No wasted energy. Just direct light output. This is why LEDs flip the efficiency ratio on its head: far more light, far less waste, and almost no heat compared to traditional bulbs.

From Christmas lights to smartphones

If you remember 1980s Christmas lights, you’ll know how unreliable and power-hungry they were. One bulb failed and the whole string went out. Modern LED versions are brighter, safer, and run on a fraction of the energy.

That same efficiency is why LEDs now power everything from TVs and smartphones to car headlights and streetlights. They’re fast, compact, energy-efficient, and can be packed into millions of pixels without overheating your screen.

Why LEDs took over the world

LEDs don’t just save energy — they last tens of thousands of hours, reduce electricity costs, and generate far less heat. In short, they’re simply better engineering.

Once you realise lighting doesn’t need to involve something dangerously close to melting point, old technology starts to feel wildly inefficient.

Want to know more? Kat, the Craig’n’Dave Lesson Hacker, explains all.

For more Lesson Hacker videos, check out the Craig’n’Dave YouTube playlist HERE.
Visit our website to explore more cutting-edge tech news in the computer science world!

Related posts

Beyond the IDE: programming lessons in the age of AI

One of the most persistent challenges in a programming classroom is knowing what every student is doing at any given […]

23 October 2026

If a NOT gate turns zero into one, where does that one come from?

How does a NOT gate turn a 0 into a 1 without creating something from nothing? Discover how voltage, transistors and digital logic work together to power the technology inside every modern computer.

15 October 2026

Ofsted after the Big Listen: have the rules changed, or just the packaging?

Ofsted has changed its inspection framework, but many school leaders question whether it has changed the culture surrounding inspection. We explore whether report cards can rebuild trust, reduce pressure and create an accountability system that genuinely helps schools improve.

9 October 2026

Should I make my own resources?

Should teachers really be making all their own resources? We explore the balance between creating, buying and adapting resources — and why thoughtful planning matters more than who made them.

5 October 2026

How does age verification technology work?

Can technology really tell how old you are? Explore the computer science behind age verification and the privacy and accuracy questions it raises.

28 September 2026

UTF-8 is witchcraft, right?

Ever wondered how your computer can happily store a simple English letter, an emoji, a pound sign and characters from […]

22 September 2026

Why teaching models keep returning

Teaching frameworks can provide valuable guidance, but when they become rigid formulas, they can undermine professional judgement. The goal isn’t identical lessons – it’s consistently effective teaching.

18 September 2026

Do GCSEs doom students to fail?

GCSE results may look stubbornly flat, but the reality behind the numbers tells a very different story. Why has further improvement become so difficult, and are we measuring success in the right way?

11 September 2026

Will Quantum Computers Break Encryption?

Could quantum computers really break encryption and put our data at risk? We explore the reality behind the hype, from qubits and Shor’s Algorithm to the future of post-quantum security.

7 September 2026

Back

Why Pixar still chooses CPUs over GPUs

29 July 2026

When Speed Isn’t Everything

At first glance, it sounds crazy. Pixar creates some of the most visually stunning films in the world — movies that could melt a graphics card for breakfast. Surely they’d be all over GPUs, right? After all, GPUs are fast, parallel, and literally designed to draw pictures. Yet, for the final render of a Pixar film, the heavy lifting is still overwhelmingly done on CPUs.
The reason isn’t speed — it’s accuracy.

GPUs vs CPUs: Games vs Film

GPUs were born in the world of gaming, where the goal is to push out 60, 120, or even more frames per second. If a shadow is slightly off in one frame, no one notices because the next frame appears in just a few milliseconds. To achieve this, GPUs use approximations: reduced-precision maths, slightly lossy floating-point operations, and other “good enough” tricks that prioritise speed over perfect accuracy.

Film rendering is a completely different challenge. Pixar’s RenderMan performs physically based rendering, tracing how light bounces, scatters, reflects, refracts, and interacts with materials — all in accordance with real-world physics. A single frame can take hours to render, not milliseconds. Any tiny calculation error is baked into the frame, ready for viewers to pause, zoom, and scrutinise on a 4K screen. That minuscule numerical wobble can lead to flickering, noise, or inconsistent lighting — all visual issues that would ruin the cinematic experience.

Why CPUs Still Rule for Final Renders

CPUs may be slower, but they are predictable. They offer high-precision floating-point maths, massive caches, and robust handling of complex branching logic — exactly what film renderers rely on. Rendering a scene isn’t just about drawing millions of pixels; it involves millions of conditional decisions about ray interactions, materials, and energy propagation. CPUs excel at this kind of irregular, decision-heavy workload, even if they take more time.

Modern GPUs are catching up and are increasingly used for previews, lighting tests, and interactive look-development, where speed matters more than perfection. But when it comes to the final frames — the ones that will live forever on the big screen — Pixar still trusts massive CPU render farms, grinding away one exquisitely accurate photon bounce at a time.

Curious to see why Pixar relies on CPUs for its blockbuster films? Watch the full video HERE.

For more fascinating insights into computer science, animation, and tech behind the scenes, visit CraignDave.org.

 

Related posts

Beyond the IDE: programming lessons in the age of AI

One of the most persistent challenges in a programming classroom is knowing what every student is doing at any given […]

23 October 2026

If a NOT gate turns zero into one, where does that one come from?

How does a NOT gate turn a 0 into a 1 without creating something from nothing? Discover how voltage, transistors and digital logic work together to power the technology inside every modern computer.

15 October 2026

Ofsted after the Big Listen: have the rules changed, or just the packaging?

Ofsted has changed its inspection framework, but many school leaders question whether it has changed the culture surrounding inspection. We explore whether report cards can rebuild trust, reduce pressure and create an accountability system that genuinely helps schools improve.

9 October 2026

Should I make my own resources?

Should teachers really be making all their own resources? We explore the balance between creating, buying and adapting resources — and why thoughtful planning matters more than who made them.

5 October 2026

How does age verification technology work?

Can technology really tell how old you are? Explore the computer science behind age verification and the privacy and accuracy questions it raises.

28 September 2026

UTF-8 is witchcraft, right?

Ever wondered how your computer can happily store a simple English letter, an emoji, a pound sign and characters from […]

22 September 2026

Why teaching models keep returning

Teaching frameworks can provide valuable guidance, but when they become rigid formulas, they can undermine professional judgement. The goal isn’t identical lessons – it’s consistently effective teaching.

18 September 2026

Do GCSEs doom students to fail?

GCSE results may look stubbornly flat, but the reality behind the numbers tells a very different story. Why has further improvement become so difficult, and are we measuring success in the right way?

11 September 2026

Will Quantum Computers Break Encryption?

Could quantum computers really break encryption and put our data at risk? We explore the reality behind the hype, from qubits and Shor’s Algorithm to the future of post-quantum security.

7 September 2026

Back

How does AI generate images?

How AI Generates Images: From noise to art

28 July 2026

What happens when you type a prompt?

AI image generation might feel like magic, but it’s actually built on surprisingly logical maths. At its core, modern systems such as diffusion models work by answering a simple question billions of times: what should this pixel probably look like?
Instead of starting with a blank canvas, AI begins with pure noise — a chaotic static image. During training, the model studies millions of real images, then gradually corrupts them with noise. Its job is to learn how to reverse that process step by step, rebuilding clarity from chaos.

From text to pixels: how prompts are understood

When you type something like “a cat wearing a top hat”, the AI doesn’t “see” words the way humans do. Instead, a text embedding model converts your prompt into a mathematical representation in a high-dimensional space. This acts like a guiding map for the image generation process.
The diffusion model then uses this guidance while cleaning up the noise. At each stage, it asks what looks incorrect and adjusts the image slightly. Over hundreds or thousands of steps, vague shapes become defined objects — whiskers, textures, lighting, and even those oddly convincing (or slightly unsettling) human features.

Why AI images look different every time

Under the hood, neural networks analyse patterns at multiple levels. Early layers detect simple shapes like edges and colour blobs, while deeper layers interpret complex structures such as eyes, fabrics, or objects.
Importantly, AI doesn’t retrieve images from a database or copy existing artwork. It generates new outputs by sampling from learned probability distributions. That’s why even identical prompts produce different results.
A controlled amount of randomness is added at the start of the process. Too little, and outputs become repetitive. Too much, and things get surreal very quickly — sometimes brilliantly so.

The real “magic” behind AI image generation

AI isn’t imagining in the human sense. It’s performing highly sophisticated statistical reconstruction, refining noise into coherent images through repeated calculation. The real breakthrough is that this process can produce visuals that feel creative, original, and often astonishingly realistic.

Want to learn more? Watch our Lesson Hacker video HERE to see exactly how AI turns text into images step by step.

 

For more Lesson Hacker videos, check out the CraignDave YouTube playlist HERE.

Visit our website to explore more cutting-edge tech news in the computer science world!

Related posts

Beyond the IDE: programming lessons in the age of AI

One of the most persistent challenges in a programming classroom is knowing what every student is doing at any given […]

23 October 2026

If a NOT gate turns zero into one, where does that one come from?

How does a NOT gate turn a 0 into a 1 without creating something from nothing? Discover how voltage, transistors and digital logic work together to power the technology inside every modern computer.

15 October 2026

Ofsted after the Big Listen: have the rules changed, or just the packaging?

Ofsted has changed its inspection framework, but many school leaders question whether it has changed the culture surrounding inspection. We explore whether report cards can rebuild trust, reduce pressure and create an accountability system that genuinely helps schools improve.

9 October 2026

Should I make my own resources?

Should teachers really be making all their own resources? We explore the balance between creating, buying and adapting resources — and why thoughtful planning matters more than who made them.

5 October 2026

How does age verification technology work?

Can technology really tell how old you are? Explore the computer science behind age verification and the privacy and accuracy questions it raises.

28 September 2026

UTF-8 is witchcraft, right?

Ever wondered how your computer can happily store a simple English letter, an emoji, a pound sign and characters from […]

22 September 2026

Why teaching models keep returning

Teaching frameworks can provide valuable guidance, but when they become rigid formulas, they can undermine professional judgement. The goal isn’t identical lessons – it’s consistently effective teaching.

18 September 2026

Do GCSEs doom students to fail?

GCSE results may look stubbornly flat, but the reality behind the numbers tells a very different story. Why has further improvement become so difficult, and are we measuring success in the right way?

11 September 2026

Will Quantum Computers Break Encryption?

Could quantum computers really break encryption and put our data at risk? We explore the reality behind the hype, from qubits and Shor’s Algorithm to the future of post-quantum security.

7 September 2026

Back

Can AI remember?

From Goldfish Memory to Game-Changing Intelligence

27 July 2026

Why does AI forget everything so quickly?

Ever felt like you’re chatting with an AI that starts off brilliantly… then completely forgets everything a few messages later?

You’re not imagining it. Most modern AI systems rely on something called a context window—a limited amount of text they can “see” at once. Once that limit is reached, older parts of the conversation simply drop off.

In simple terms, AI doesn’t “remember” in the way humans do. Each time you send a message, the system rereads what fits within its token limit. Anything beyond that? Gone. It’s like revising for an exam on a whiteboard that keeps erasing your first notes as you add new ones.

The real challenge: scaling memory and attention

The issue isn’t just memory—it’s also performance. The attention mechanism that powers AI becomes increasingly expensive as conversations grow longer. Double the input, and the computational cost can skyrocket. That’s why today’s AI works well for short tasks, but struggles with long, complex interactions.

A smarter future: short-term vs long-term AI memory

Exciting new developments are aiming to fix this. Emerging approaches—like Google’s “Titans” architecture—introduce the idea of splitting AI memory into two parts:

  • Short-term memory for immediate context (like current conversations).
  • Long-term memory for storing important information over time.

This long-term memory works more like a structured system, allowing AI to retain key details without cramming everything into a limited space.

The “surprise” factor: how AI decides what to remember

One of the most fascinating ideas is the use of a surprise metric. Instead of remembering everything, AI prioritises unusual or important information—just like humans do.

Think of it as a bouncer for memory:

  • Predictable, repetitive information gets filtered out.
  • Unexpected or meaningful details get stored long-term.

This makes AI far more efficient and helps it retain what actually matters.

What this means in the real world

If these advancements continue, we could see:

  • Coding assistants that remember your entire project history
  • Study tools that retain knowledge across the academic year
  • Games with characters that build genuine long-term relationships
  • Customer support AI that actually remembers you (no more repeating yourself!)

In short, AI could become more personal, consistent, and genuinely useful over time.

But what about privacy?

With great memory comes great responsibility. Systems that store long-term information will need clear controls—what gets remembered, what gets forgotten, and why. Without this, there’s a real risk of storing incorrect or sensitive data.

We’re moving towards AI that doesn’t just respond—but remembers. Less goldfish, more elephant (a very fast one). And that shift could completely transform how we interact with technology.

Want to see this explained in action? Watch our Lesson Hacker video on our YouTube channel.

For more Lesson Hacker videos, check out the CraignDave YouTube playlist HERE.

Visit our website to explore more cutting-edge tech news in the computer science world!

 

Related posts

Beyond the IDE: programming lessons in the age of AI

One of the most persistent challenges in a programming classroom is knowing what every student is doing at any given […]

23 October 2026

If a NOT gate turns zero into one, where does that one come from?

How does a NOT gate turn a 0 into a 1 without creating something from nothing? Discover how voltage, transistors and digital logic work together to power the technology inside every modern computer.

15 October 2026

Ofsted after the Big Listen: have the rules changed, or just the packaging?

Ofsted has changed its inspection framework, but many school leaders question whether it has changed the culture surrounding inspection. We explore whether report cards can rebuild trust, reduce pressure and create an accountability system that genuinely helps schools improve.

9 October 2026

Should I make my own resources?

Should teachers really be making all their own resources? We explore the balance between creating, buying and adapting resources — and why thoughtful planning matters more than who made them.

5 October 2026

How does age verification technology work?

Can technology really tell how old you are? Explore the computer science behind age verification and the privacy and accuracy questions it raises.

28 September 2026

UTF-8 is witchcraft, right?

Ever wondered how your computer can happily store a simple English letter, an emoji, a pound sign and characters from […]

22 September 2026

Why teaching models keep returning

Teaching frameworks can provide valuable guidance, but when they become rigid formulas, they can undermine professional judgement. The goal isn’t identical lessons – it’s consistently effective teaching.

18 September 2026

Do GCSEs doom students to fail?

GCSE results may look stubbornly flat, but the reality behind the numbers tells a very different story. Why has further improvement become so difficult, and are we measuring success in the right way?

11 September 2026

Will Quantum Computers Break Encryption?

Could quantum computers really break encryption and put our data at risk? We explore the reality behind the hype, from qubits and Shor’s Algorithm to the future of post-quantum security.

7 September 2026

Back

How do drones keep balanced?

The science behind hovering magic

23 July 2026

Ever watched a drone hover perfectly in mid-air and wondered how it doesn’t just tip over like a table on uneven ground? The truth is fascinating — drones don’t balance by being still. They balance by panicking hundreds of times a second… and doing it with incredible precision.

Tiny Sensors, Big Job

A quadcopter isn’t stable on its own. Left alone, it would immediately tip, accelerate sideways, and fall — obeying gravity in the most dramatic way. The secret lies inside: an Inertial Measurement Unit (IMU). This tiny device, packed with accelerometers and gyroscopes, measures linear acceleration and rotation across three axes. Essentially, the IMU is constantly asking, “Which way is down, and how fast are we messing this up?”

The Control Loop: Chaos with a Plan

The IMU feeds its measurements into a flight controller, usually using a PID controller — Proportional, Integral, Derivative.

This may sound technical, but it’s simply a way to answer three questions:

  • How wrong are we now?
  • How wrong have we been?
  • How fast is it getting worse?

If the drone tilts even slightly forward, the controller speeds up the rear motors and slows the front ones. This tiny adjustment pushes it level again. And it doesn’t stop there — this process happens hundreds or thousands of times per second, far faster than any human pilot could react.

Torque Cancellation: Why Multiple Rotors Matter

Drones have multiple rotors to cancel out torque. Each spinning propeller wants to twist the drone in the opposite direction (thanks, Newton!). By pairing clockwise and counter-clockwise rotors, rotational forces cancel each other out. Want to yaw left? Speed up one diagonal pair and slow the other. Want to rise? Speed up all the motors equally.

Advanced Awareness: More Than Just Balance

GPS, barometers, magnetometers, and even vision systems give drones higher-level awareness — altitude, heading, and position relative to the ground. But these sensors are slower. The real magic of staying upright happens in the IMU and control loop: a relentless feedback cycle of overcorrecting tiny mistakes instantly.

Drones don’t hover by being steady — they hover by constantly falling and correcting faster than physics can react. What looks like effortless floating is actually hundreds of tiny “NOPE!” corrections every second. The result? A perfectly balanced drone and a glimpse into the brilliant engineering of modern flight.

Watch the full video HERE to see this balancing act in action.

Explore more tech insights and computer science fun at CraignDave.org

Related posts

Beyond the IDE: programming lessons in the age of AI

One of the most persistent challenges in a programming classroom is knowing what every student is doing at any given […]

23 October 2026

If a NOT gate turns zero into one, where does that one come from?

How does a NOT gate turn a 0 into a 1 without creating something from nothing? Discover how voltage, transistors and digital logic work together to power the technology inside every modern computer.

15 October 2026

Ofsted after the Big Listen: have the rules changed, or just the packaging?

Ofsted has changed its inspection framework, but many school leaders question whether it has changed the culture surrounding inspection. We explore whether report cards can rebuild trust, reduce pressure and create an accountability system that genuinely helps schools improve.

9 October 2026

Should I make my own resources?

Should teachers really be making all their own resources? We explore the balance between creating, buying and adapting resources — and why thoughtful planning matters more than who made them.

5 October 2026

How does age verification technology work?

Can technology really tell how old you are? Explore the computer science behind age verification and the privacy and accuracy questions it raises.

28 September 2026

UTF-8 is witchcraft, right?

Ever wondered how your computer can happily store a simple English letter, an emoji, a pound sign and characters from […]

22 September 2026

Why teaching models keep returning

Teaching frameworks can provide valuable guidance, but when they become rigid formulas, they can undermine professional judgement. The goal isn’t identical lessons – it’s consistently effective teaching.

18 September 2026

Do GCSEs doom students to fail?

GCSE results may look stubbornly flat, but the reality behind the numbers tells a very different story. Why has further improvement become so difficult, and are we measuring success in the right way?

11 September 2026

Will Quantum Computers Break Encryption?

Could quantum computers really break encryption and put our data at risk? We explore the reality behind the hype, from qubits and Shor’s Algorithm to the future of post-quantum security.

7 September 2026

Back

Is fetch, decode, execute still a thing?

22 July 2026

Yes, it is—but it’s a lot more complicated than you remember.

If you remember your early computer science lessons, you might recall the simple mantra: fetch an instruction, decode it, execute it, repeat. That basic loop has been at the heart of CPUs since the 1970s—and surprisingly, it’s still there today. But modern processors have bulked up, gotten faster, and learned a few tricks along the way.

The classic rhythm of CPUs

At its core, every CPU still follows the same three-step process: fetch an instruction from memory, decode it into an operation the processor understands, and execute it. Software assumes this rhythm, and physics hasn’t suggested a better alternative yet. Strip away the cores, branding, and flashy diagrams, and the familiar fetch–decode–execute cycle remains.

Pipelining: the relay race of instructions

What’s different now is speed and efficiency. Modern CPUs don’t sit politely and handle one instruction at a time. Instead, they fetch multiple instructions ahead, decode them in parallel, and try to execute as many as possible simultaneously. This is called pipelining. Picture a relay race where runners start early, pass the baton multiple times at once, and occasionally bump into each other—chaotic but effective.

Out-of-order execution and speculation

Things get even more “clever” (and a little chaotic) with out-of-order execution. CPUs don’t always follow the order your program wrote; they execute instructions as resources are available, as long as the final result is correct. Add branch prediction and speculative execution, and your processor is effectively guessing what comes next, executing instructions in advance, and discarding the ones it didn’t need. If done right, this adds speed. If done wrong, well… let’s just say it occasionally makes headlines.

The same old cycle, just supercharged

Underneath all the complexity, the fundamentals haven’t disappeared. Instructions are still fetched, decoded, and executed. Even GPUs, AI accelerators, and multi-core monsters rely on this cycle—they’ve just widened, deepened, parallelised, and turbocharged it. Today’s CPU is like an over-caffeinated octopus juggling thousands of tasks at once—but from the software’s perspective, it still looks like a simple three-step dance.

So yes, fetch, decode, execute is absolutely still a thing. It’s just gone to the gym, had some coffee, and learned a few new moves along the way.

Want to see this explained visually? Check out the full video on our CraignDave YouTube channel.


Explore more computer science insights, tutorials, and resources at CraignDave.

Related posts

Beyond the IDE: programming lessons in the age of AI

One of the most persistent challenges in a programming classroom is knowing what every student is doing at any given […]

23 October 2026

If a NOT gate turns zero into one, where does that one come from?

How does a NOT gate turn a 0 into a 1 without creating something from nothing? Discover how voltage, transistors and digital logic work together to power the technology inside every modern computer.

15 October 2026

Ofsted after the Big Listen: have the rules changed, or just the packaging?

Ofsted has changed its inspection framework, but many school leaders question whether it has changed the culture surrounding inspection. We explore whether report cards can rebuild trust, reduce pressure and create an accountability system that genuinely helps schools improve.

9 October 2026

Should I make my own resources?

Should teachers really be making all their own resources? We explore the balance between creating, buying and adapting resources — and why thoughtful planning matters more than who made them.

5 October 2026

How does age verification technology work?

Can technology really tell how old you are? Explore the computer science behind age verification and the privacy and accuracy questions it raises.

28 September 2026

UTF-8 is witchcraft, right?

Ever wondered how your computer can happily store a simple English letter, an emoji, a pound sign and characters from […]

22 September 2026

Why teaching models keep returning

Teaching frameworks can provide valuable guidance, but when they become rigid formulas, they can undermine professional judgement. The goal isn’t identical lessons – it’s consistently effective teaching.

18 September 2026

Do GCSEs doom students to fail?

GCSE results may look stubbornly flat, but the reality behind the numbers tells a very different story. Why has further improvement become so difficult, and are we measuring success in the right way?

11 September 2026

Will Quantum Computers Break Encryption?

Could quantum computers really break encryption and put our data at risk? We explore the reality behind the hype, from qubits and Shor’s Algorithm to the future of post-quantum security.

7 September 2026

Back

Why is RAM Suddenly So Expensive?

21 July 2026

It feels like just yesterday that RAM upgrades were practically free – and now a 32 GB stick can cost almost as much as a short city break. If you’ve been scratching your head over soaring memory prices, the usual explanation you’ll hear is “AI.” It sounds suspiciously vague… but in this case, it’s annoyingly accurate – just not in the way most people assume.

The Memory Family: RAM, VRAM, and HBM

To understand what’s happening, you need to know a bit about memory. RAM, VRAM, and the ultra-fancy AI memory called HBM are essentially cousins. All are forms of DRAM, manufactured using very similar processes, in the same factories, and by a surprisingly small number of companies – namely Samsung, SK Hynix, and Micron. That’s it. Three firms producing the vast majority of the world’s memory.
Think of it less like a bustling free market and more like three bakers deciding what bread exists in town.

How AI Data Centres Are Hoovering Memory

Enter AI data centres. Training and running large models isn’t just about GPUs packed with VRAM – they also need enormous amounts of system RAM to shuffle data, stage batches, and keep everything moving smoothly. A single AI server can easily house terabytes of DDR5 RAM, and a single data centre can contain thousands of such servers. That’s not “a bit more demand” – that’s hoovering the warehouse.

Why Memory Prices Are Climbing

Here’s the catch: memory fabs can’t just whip up more chips overnight. Building or expanding a fabrication plant takes years and costs billions. Memory is famously a boom-and-bust industry, and manufacturers have been burned before by ramping up production right before demand collapsed.
So, instead of expanding, they’re making the rational move: prioritising the most profitable memory. Right now, that’s HBM for AI accelerators, which sell for far higher margins than boring old consumer DDR5.
The result? Less capacity goes to consumer RAM, more goes to high-margin AI memory, and prices climb. Your PC isn’t suddenly demanding AI – it’s just that data centres can outbid you without even noticing you were in the room.

The Takeaway

Your next RAM upgrade isn’t expensive because it’s better. It’s expensive because someone else, somewhere, is willing to pay vastly more. Understanding the forces behind memory pricing helps make sense of those shocking sticker prices – and shows just how intertwined AI and everyday computing have become.

Watch the full video to see the story behind the RAM price hike in action.

Explore more computer science insights and resources at CraignDave.org

Related posts

Beyond the IDE: programming lessons in the age of AI

One of the most persistent challenges in a programming classroom is knowing what every student is doing at any given […]

23 October 2026

If a NOT gate turns zero into one, where does that one come from?

How does a NOT gate turn a 0 into a 1 without creating something from nothing? Discover how voltage, transistors and digital logic work together to power the technology inside every modern computer.

15 October 2026

Ofsted after the Big Listen: have the rules changed, or just the packaging?

Ofsted has changed its inspection framework, but many school leaders question whether it has changed the culture surrounding inspection. We explore whether report cards can rebuild trust, reduce pressure and create an accountability system that genuinely helps schools improve.

9 October 2026

Should I make my own resources?

Should teachers really be making all their own resources? We explore the balance between creating, buying and adapting resources — and why thoughtful planning matters more than who made them.

5 October 2026

How does age verification technology work?

Can technology really tell how old you are? Explore the computer science behind age verification and the privacy and accuracy questions it raises.

28 September 2026

UTF-8 is witchcraft, right?

Ever wondered how your computer can happily store a simple English letter, an emoji, a pound sign and characters from […]

22 September 2026

Why teaching models keep returning

Teaching frameworks can provide valuable guidance, but when they become rigid formulas, they can undermine professional judgement. The goal isn’t identical lessons – it’s consistently effective teaching.

18 September 2026

Do GCSEs doom students to fail?

GCSE results may look stubbornly flat, but the reality behind the numbers tells a very different story. Why has further improvement become so difficult, and are we measuring success in the right way?

11 September 2026

Will Quantum Computers Break Encryption?

Could quantum computers really break encryption and put our data at risk? We explore the reality behind the hype, from qubits and Shor’s Algorithm to the future of post-quantum security.

7 September 2026

Back

Am I really Human?

How CAPTCHAs & AI test your humanity

20 July 2026

The tiny checkbox with a big job

Every time you tick the familiar “I am not a robot” box online, it might feel like a small insult to your dignity. Surely robots can click boxes — they can drive cars and even compose music. So why does a simple checkbox seem to challenge our very existence?
The truth is, that little square isn’t testing whether you can click — it’s observing how you click. Humans are delightfully messy: we overshoot targets, pause, wiggle the mouse, and sometimes hesitate for a moment of existential dread before finally hitting the box. Bots, on the other hand, move in perfectly straight lines at superhuman speed, trying to complete the task with robotic efficiency. Invisible scripts analyse cursor paths, timings, micro-hesitations, and device behaviours. If the interaction looks too perfect, the system thinks: “Robot vibes,” and you’re sent to the dreaded image test.

Enter CAPTCHA hell

CAPTCHAs — those grids of blurry images — are actually computer vision training disguised as security. You’re asked to select traffic lights, bicycles, or buses that appear half-hidden or distorted. Humans excel at recognising patterns in chaos, while early AI struggled. The system even compares your selections with those of other humans, effectively crowd-sourcing a reality check. Meanwhile, you debate whether that tandem bike counts as one or two.

Can AI beat CAPTCHAs?

Increasingly, yes. Modern neural networks can identify objects in CAPTCHAs far more accurately than humans, especially when fed enough examples. This is why newer systems have evolved beyond image puzzles, relying on behavioural biometrics, device fingerprints, risk scoring, and contextual signals like your IP reputation. Proving you’re human is no longer just about ticking a box — it’s about showing you’re a normal human doing normal human things on a normal device.

Humans are gloriously clumsy

Ironically, as AI gets better at mimicking us, websites may start testing behaviours that even humans find tricky. In the future, you might need to scroll in a slightly confused way or fail a simple maths problem to prove you’re human. Finally, our natural human quirks become a security feature — and one that bots can’t easily replicate.

Curious to see how this all works in practice and why your mouse movements matter more than you think? Watch the full video here.

For more fascinating computer science insights, teaching resources, and videos, visit CraignDave.

Related posts

Beyond the IDE: programming lessons in the age of AI

One of the most persistent challenges in a programming classroom is knowing what every student is doing at any given […]

23 October 2026

If a NOT gate turns zero into one, where does that one come from?

How does a NOT gate turn a 0 into a 1 without creating something from nothing? Discover how voltage, transistors and digital logic work together to power the technology inside every modern computer.

15 October 2026

Ofsted after the Big Listen: have the rules changed, or just the packaging?

Ofsted has changed its inspection framework, but many school leaders question whether it has changed the culture surrounding inspection. We explore whether report cards can rebuild trust, reduce pressure and create an accountability system that genuinely helps schools improve.

9 October 2026

Should I make my own resources?

Should teachers really be making all their own resources? We explore the balance between creating, buying and adapting resources — and why thoughtful planning matters more than who made them.

5 October 2026

How does age verification technology work?

Can technology really tell how old you are? Explore the computer science behind age verification and the privacy and accuracy questions it raises.

28 September 2026

UTF-8 is witchcraft, right?

Ever wondered how your computer can happily store a simple English letter, an emoji, a pound sign and characters from […]

22 September 2026

Why teaching models keep returning

Teaching frameworks can provide valuable guidance, but when they become rigid formulas, they can undermine professional judgement. The goal isn’t identical lessons – it’s consistently effective teaching.

18 September 2026

Do GCSEs doom students to fail?

GCSE results may look stubbornly flat, but the reality behind the numbers tells a very different story. Why has further improvement become so difficult, and are we measuring success in the right way?

11 September 2026

Will Quantum Computers Break Encryption?

Could quantum computers really break encryption and put our data at risk? We explore the reality behind the hype, from qubits and Shor’s Algorithm to the future of post-quantum security.

7 September 2026

Back

Why are TVs so cheap now?

10 July 2026

The secret behind your £120 4K screen

Remember when buying a TV was a major event? A 32-inch telly back in the day was basically a family heirloom. You’d gather around it like it was the village fire, watching BBC One through a haze of static so thick it felt like the actors were performing behind frosted glass. And the price? A thousand pounds for a good one.
Fast forward to today, and a 43-inch 4K smart TV can cost less than a weekly meal deal. Witchcraft? Not quite. Mostly economics… and a little clever tech.

How modern TVs are made

The biggest change comes from how screens are manufactured. Old TVs relied on cathode ray tubes—massive, heavy boxes firing electrons at phosphor pixels, like some radioactive Nerf gun. Modern LCD and OLED panels, by contrast, are made with industrial precision.
Major manufacturers like BOE, AUO, LG Display, and Samsung Display produce enormous “mother glass” sheets the size of garage doors. These sheets are then sliced into individual panels like display lasagne. Larger sheets plus better yield rates dramatically lower the cost per screen.
Automation has also transformed the process. Making a TV panel today is closer to printing a newspaper than hand-crafting a delicate scientific artefact. Mass production crushes the price, plain and simple.

Why you’re part of the price equation

Here’s the spicy bit: your TV is cheap because you’re the product. Smart TVs collect data on what you watch, when you watch, and even how long you stick with Love Island before rage-quitting. This process, called Automatic Content Recognition (ACR), tracks content across apps and external devices. The data is anonymised but extremely valuable to advertisers.
In other words, manufacturers can subsidise the hardware cost because they’re making money on the software side. Think of it as a razor-and-blades model—except the razor is a 43” slab of glass, and the blades are your viewing habits.

The advertising ecosystem

Beyond data collection, your TV is prime real estate for ads and paid placements. Budget brands often earn more after you buy the TV than from the sale itself. It’s the same trick as cheap printers—but without the emotional abuse of running out of ink mid-document.

Market forces & competition

Finally, competition keeps prices low. There are countless TV manufacturers, and once you’ve nailed 4K LCD production, components like processors, LEDs, and Wi-Fi chips are cheap and commoditised. Seasonal spikes in TV purchases allow retailers to sell screens as loss-leaders, tempting you to buy a toaster, soundbar, or lifetime supply of discount stuffing along the way.

So yes, TVs are cheap because manufacturing is efficient, components are inexpensive, and companies quietly gather data about your viewing habits. That £120 miracle screen? Enjoy it… and maybe take a few minutes to disable tracking, or at least make sure your TV knows you have excellent taste.

Watch our Lesson Hacker video to find out more.

Discover more computer science insights and fun explanations at CraignDave.org.

Related posts

Beyond the IDE: programming lessons in the age of AI

One of the most persistent challenges in a programming classroom is knowing what every student is doing at any given […]

23 October 2026

If a NOT gate turns zero into one, where does that one come from?

How does a NOT gate turn a 0 into a 1 without creating something from nothing? Discover how voltage, transistors and digital logic work together to power the technology inside every modern computer.

15 October 2026

Ofsted after the Big Listen: have the rules changed, or just the packaging?

Ofsted has changed its inspection framework, but many school leaders question whether it has changed the culture surrounding inspection. We explore whether report cards can rebuild trust, reduce pressure and create an accountability system that genuinely helps schools improve.

9 October 2026

Should I make my own resources?

Should teachers really be making all their own resources? We explore the balance between creating, buying and adapting resources — and why thoughtful planning matters more than who made them.

5 October 2026

How does age verification technology work?

Can technology really tell how old you are? Explore the computer science behind age verification and the privacy and accuracy questions it raises.

28 September 2026

UTF-8 is witchcraft, right?

Ever wondered how your computer can happily store a simple English letter, an emoji, a pound sign and characters from […]

22 September 2026

Why teaching models keep returning

Teaching frameworks can provide valuable guidance, but when they become rigid formulas, they can undermine professional judgement. The goal isn’t identical lessons – it’s consistently effective teaching.

18 September 2026

Do GCSEs doom students to fail?

GCSE results may look stubbornly flat, but the reality behind the numbers tells a very different story. Why has further improvement become so difficult, and are we measuring success in the right way?

11 September 2026

Will Quantum Computers Break Encryption?

Could quantum computers really break encryption and put our data at risk? We explore the reality behind the hype, from qubits and Shor’s Algorithm to the future of post-quantum security.

7 September 2026

Back

The Year 7 dip: why enthusiasm fades – and how to rebuild it

7 July 2026

Every secondary teacher knows the feeling. A Year 7 class arrives in September full of enthusiasm — hands up, eyes forward, eager to please. Fast forward a few months, and the same class can feel markedly different: quieter, less responsive, occasionally resistant.

Writing in TES, David Thomas, CEO of Axiom Maths, describes this familiar experience. His research suggests that, by the summer of Year 7, students are significantly less likely to enjoy Maths than they were in Year 6. What is striking is not just that this happens, but how quickly — and how unevenly.

Yet as many classroom teachers will recognise, this pattern is not confined to Maths. It can affect Computing too, albeit playing out on a slightly different timeline.

When the shine wears off

Thomas identifies the spring term, between Christmas and Easter as the critical point where engagement drops most sharply. With multiple lessons a week, Maths provides fertile ground for this decline to emerge quickly.

In Computing, Craig and Dave suggest the change is often delayed. With fewer curriculum hours, students tend to sustain their enthusiasm through Year 7. But the dip still comes.

Year sevens are generally keen all year, but it’s when they come back in Year 8 that you really notice the shift. They’re more comfortable in the school and more complacent about learning.

This is a useful reminder, the issue is not tied to a specific term, but to a broader transition. As students settle into secondary school, the initial novelty fades and deeper challenges begin to surface.

The balancing act teachers know too well

One of Thomas’s key findings centres on the tension between repetition and pace. Teachers must reconcile wildly different starting points, often reteaching content to establish a common foundation.

In principle, this is sound pedagogy. In practice, it can be deeply frustrating for students with some thinking, “We’ve already done this — why are we still here?” Others are completely bamboozled and trying to keep up.

In Computing, this disparity can feel even more pronounced. Pupils arrive with highly variable experiences — from those who have explored coding in depth to those who have barely used a keyboard. The result is a classroom where boredom and anxiety coexist.

The danger lies in assuming that this tension is unavoidable. While it may not be fully solvable, thoughtful planning, adaptive tasks, extension pathways, low-threshold/high-ceiling activities can mitigate its impact.

 

Confidence: the slow erosion

If the mismatch of difficulty is the spark, then loss of confidence is the slow-burning fire. Thomas’s research highlights a growing confidence gap — one that emerges before any clear attainment gap. Teachers see this every day, though often in subtle ways. In Year 7, when you pose a question to the class, it’s “me, me, me” — hands shooting up. By Year 8, some students just… stop. They don’t want to put themselves out there anymore.

What changes is not just academic confidence, but social awareness. Participation becomes risky. It is no longer about impressing the teacher, but about fitting in with peers.

This creates a difficult tension. The very behaviours that support learning — asking questions, making mistakes, persevering — are precisely those that can feel socially uncomfortable. Left unchecked, this quiet withdrawal becomes self-reinforcing. The more you think students don’t want to engage, the less you push it, and the less you push it, the less they do. It becomes a downward spiral.

Classroom climate: more than behaviour management

Thomas also identifies a decline in classroom climate — reduced collaboration, less effective listening, and increased disruption. This is not simply a behaviour issue, but a cultural one. Being “too cool for school” changes everything.

Once that shift takes hold, even well-planned lessons struggle to land. Group work becomes harder to sustain. Discussion loses its richness. Students who want to engage can feel inhibited by those around them. However, sustaining high expectations is essential, even when it feels difficult. The moment you lose heart, they lose heart.

The unequal recovery

Perhaps the most concerning aspect of Thomas’s findings is what happens next. Some students begin to re-engage — but this recovery is not evenly distributed. More advantaged pupils are more likely to regain enjoyment. Others remain disengaged, widening the gap in both attitude and, eventually, attainment.

Teachers recognise this pattern, even if they do not always name it explicitly. Some pupils find their way back through extracurricular activities or additional support. Others drift further away. For a long time, both teachers and students have responded to this by asking whether the curriculum is truly fit for purpose for all learners.

This tension sits at the heart of inclusive teaching. The goal is not universal passion, but universal opportunity.

Beyond “teenagers being teenagers”

A common explanation for declining motivation is adolescence itself. Hormones, identity, social pressures all play a role, but we should caution against using it as a catch-all explanation. Instead, perhaps there is a deeper mismatch between student needs and school structures.

As adolescents develop, they increasingly seek:

  • autonomy — choice and independence;
  • competence — a sense of progress and success;
  • belonging — connection and identity.

Yet, secondary schools can often drift in the opposite direction creating environments that are controlling, performance-focused which results in spoon-feeding and impersonal. Then we’re surprised when motivation drops.

This is not an indictment of individuals, but of systems. Teachers operate within demanding structures — large classes, heavy workloads, constant pressures. It is easy to fall into cycles of survival rather than reflection. But awareness is a starting point.

Rebuilding the spark

The motivation dip is not inevitable. Nor is it irreversible. If anything, the research and classroom reflections suggest that there is a crucial window — particularly towards the end of Year 7 where small, deliberate shifts can have lasting impact.

Some principles stand out:

  • Protect confidence early
    Notice the students who have gone quiet. Create opportunities where success feels visible and safe.
  • Guard the classroom culture
    Maintain high expectations for participation and collaboration, even when it is challenging.
  • Keep the subject ‘alive’
    Don’t strip out engaging activities simply because they are harder to manage.
  • Offer meaningful challenge
    As Thomas argues, the answer is not easier content, but richer experiences — problems that genuinely engage curiosity.

A profession worth backing

Ultimately, this is as much about teachers as it is about students. Sustaining engagement requires energy, reflection and, at times, resilience. Teaching is not a theoretical exercise. It is lived, complex, and often messy, but within that complexity lies opportunity. The patterns we see are not fixed; they are shaped by the environments we create. While no classroom is perfect, every classroom has the potential to shift the trajectory. The Year 7 dip may be predictable, but it is also preventable.

Want to know more? Check out our ‘At the chalk face’ episode on YouTube.

Related posts

Beyond the IDE: programming lessons in the age of AI

One of the most persistent challenges in a programming classroom is knowing what every student is doing at any given […]

23 October 2026

If a NOT gate turns zero into one, where does that one come from?

How does a NOT gate turn a 0 into a 1 without creating something from nothing? Discover how voltage, transistors and digital logic work together to power the technology inside every modern computer.

15 October 2026

Ofsted after the Big Listen: have the rules changed, or just the packaging?

Ofsted has changed its inspection framework, but many school leaders question whether it has changed the culture surrounding inspection. We explore whether report cards can rebuild trust, reduce pressure and create an accountability system that genuinely helps schools improve.

9 October 2026

Should I make my own resources?

Should teachers really be making all their own resources? We explore the balance between creating, buying and adapting resources — and why thoughtful planning matters more than who made them.

5 October 2026

How does age verification technology work?

Can technology really tell how old you are? Explore the computer science behind age verification and the privacy and accuracy questions it raises.

28 September 2026

UTF-8 is witchcraft, right?

Ever wondered how your computer can happily store a simple English letter, an emoji, a pound sign and characters from […]

22 September 2026

Why teaching models keep returning

Teaching frameworks can provide valuable guidance, but when they become rigid formulas, they can undermine professional judgement. The goal isn’t identical lessons – it’s consistently effective teaching.

18 September 2026

Do GCSEs doom students to fail?

GCSE results may look stubbornly flat, but the reality behind the numbers tells a very different story. Why has further improvement become so difficult, and are we measuring success in the right way?

11 September 2026

Will Quantum Computers Break Encryption?

Could quantum computers really break encryption and put our data at risk? We explore the reality behind the hype, from qubits and Shor’s Algorithm to the future of post-quantum security.

7 September 2026

Back

Will quantum computers break encryption?

6 July 2026

The Truth Behind the Hype

Quantum computers are all over the headlines, and one question keeps cyber-security experts — and curious Reddit users — awake at night: will they break encryption? Let’s unpack this complex topic in a way that’s informative, a little fun, and very relevant for computer science enthusiasts.

What makes quantum computers different?

A standard computer thinks in bits — ones and zeroes, on or off. Quantum computers, however, use qubits, which can exist as one, zero, and everything in between simultaneously. This phenomenon, called superposition, allows quantum computers to explore multiple possibilities at once. Imagine trying every combination of a password simultaneously — it’s like a student’s excuse for not doing homework, but way more powerful.

Shor’s Algorithm: Why experts worry

The panic around quantum computing often centres on Shor’s Algorithm. This quantum method can factor extremely large numbers at lightning speed. Most modern encryption, such as RSA, relies on the difficulty of breaking these numbers into prime factors — essentially trying to un-mix a smoothie. With Shor’s Algorithm, those numbers could be split back into their ingredients in seconds. If a sufficiently powerful quantum computer existed, today’s encryption could be compromised.

The reality check: we’re not there yet

Here’s the reassuring bit: current quantum computers have only a few hundred noisy qubits — prone to mistakes, like a Year 10 coding project. To break modern encryption, we’d need millions of perfectly stable qubits with near-perfect error correction. This is decades away, requires extreme cooling near absolute zero, and a level of engineering that makes the Large Hadron Collider look like a Lego set.

Post-quantum encryption: staying ahead of the curve

By the time quantum computers are powerful enough to threaten encryption, post-quantum encryption will likely be in place. These new cryptographic methods are designed to withstand quantum attacks, keeping your data safe. In short, quantum computing is exciting — but not yet a cyber apocalypse.

Until then, your passwords are safe — unless you’re still using “password123”. In that case, your cat on the keyboard is a bigger threat.

Want to learn more about quantum computing and cyber security? Watch the full explainer video HERE.

Explore more computer science insights, lesson plans, and resources at craigndave.org

Related posts

Beyond the IDE: programming lessons in the age of AI

One of the most persistent challenges in a programming classroom is knowing what every student is doing at any given […]

23 October 2026

If a NOT gate turns zero into one, where does that one come from?

How does a NOT gate turn a 0 into a 1 without creating something from nothing? Discover how voltage, transistors and digital logic work together to power the technology inside every modern computer.

15 October 2026

Ofsted after the Big Listen: have the rules changed, or just the packaging?

Ofsted has changed its inspection framework, but many school leaders question whether it has changed the culture surrounding inspection. We explore whether report cards can rebuild trust, reduce pressure and create an accountability system that genuinely helps schools improve.

9 October 2026

Should I make my own resources?

Should teachers really be making all their own resources? We explore the balance between creating, buying and adapting resources — and why thoughtful planning matters more than who made them.

5 October 2026

How does age verification technology work?

Can technology really tell how old you are? Explore the computer science behind age verification and the privacy and accuracy questions it raises.

28 September 2026

UTF-8 is witchcraft, right?

Ever wondered how your computer can happily store a simple English letter, an emoji, a pound sign and characters from […]

22 September 2026

Why teaching models keep returning

Teaching frameworks can provide valuable guidance, but when they become rigid formulas, they can undermine professional judgement. The goal isn’t identical lessons – it’s consistently effective teaching.

18 September 2026

Do GCSEs doom students to fail?

GCSE results may look stubbornly flat, but the reality behind the numbers tells a very different story. Why has further improvement become so difficult, and are we measuring success in the right way?

11 September 2026

Will Quantum Computers Break Encryption?

Could quantum computers really break encryption and put our data at risk? We explore the reality behind the hype, from qubits and Shor’s Algorithm to the future of post-quantum security.

7 September 2026

Back

How does that website know where I live?

5 July 2026

The spooky “Good evening, Manchester!” moment

You’ve opened a website, and it greets you with “Good evening, Manchester!” and your first thought is: Right. Which one of you told them? Don’t worry—it’s not psychic powers, and your phone isn’t secretly spying on you. Unless you’ve been oversharing every thought on social media, it’s probably just IP geolocation—the tech equivalent of guessing someone’s postcode from the accent they mumble through a kebab.

IP addresses: Your digital return address

Every device connecting to the internet gets an IP address, a numeric label that works a bit like a return address on a letter (less charming, more colons if it’s IPv6). Internet providers assign these addresses in big regional blocks, and companies like MaxMind and IP2Location gather this data into huge databases mapping IP ranges to countries, cities, and sometimes even neighbourhoods.
So when you visit a website, it checks your IP against one of these databases and voilà—“Ah, this one’s from Leeds.” Not exactly Sherlock-level deduction, but surprisingly effective.

Why it’s not always perfect

IP geolocation isn’t 100% precise. Your village might be resolved to the nearest city, or worse, a default entry could place you in a random field in Kansas! Bad data can cause chaos—but it’s usually harmless.
And if you’re worried about your exact house number, rest easy. Websites would need your ISP’s internal records to get that, which are tightly guarded. Only a court order could lift that veil.

When accuracy improves

IP geolocation can be more precise if you use apps or services that access GPS, Wi-Fi positioning, or mobile networks—but only if you explicitly share that data. Otherwise, a plain IP lookup just gives a rough “local-ish vibe.”

Next time a website welcomes you by your city, don’t panic. It’s not magic, it’s not sinister surveillance, and there’s no crystal ball involved. It’s just your internet address doing the talking.
Want to know more? Watch our Lesson Hacker video on our YouTube channel.

For more Lesson Hacker videos, check out the CraignDave YouTube playlist HERE.

Visit our website to explore more cutting-edge tech news in the computer science world!

Related posts

Beyond the IDE: programming lessons in the age of AI

One of the most persistent challenges in a programming classroom is knowing what every student is doing at any given […]

23 October 2026

If a NOT gate turns zero into one, where does that one come from?

How does a NOT gate turn a 0 into a 1 without creating something from nothing? Discover how voltage, transistors and digital logic work together to power the technology inside every modern computer.

15 October 2026

Ofsted after the Big Listen: have the rules changed, or just the packaging?

Ofsted has changed its inspection framework, but many school leaders question whether it has changed the culture surrounding inspection. We explore whether report cards can rebuild trust, reduce pressure and create an accountability system that genuinely helps schools improve.

9 October 2026

Should I make my own resources?

Should teachers really be making all their own resources? We explore the balance between creating, buying and adapting resources — and why thoughtful planning matters more than who made them.

5 October 2026

How does age verification technology work?

Can technology really tell how old you are? Explore the computer science behind age verification and the privacy and accuracy questions it raises.

28 September 2026

UTF-8 is witchcraft, right?

Ever wondered how your computer can happily store a simple English letter, an emoji, a pound sign and characters from […]

22 September 2026

Why teaching models keep returning

Teaching frameworks can provide valuable guidance, but when they become rigid formulas, they can undermine professional judgement. The goal isn’t identical lessons – it’s consistently effective teaching.

18 September 2026

Do GCSEs doom students to fail?

GCSE results may look stubbornly flat, but the reality behind the numbers tells a very different story. Why has further improvement become so difficult, and are we measuring success in the right way?

11 September 2026

Will Quantum Computers Break Encryption?

Could quantum computers really break encryption and put our data at risk? We explore the reality behind the hype, from qubits and Shor’s Algorithm to the future of post-quantum security.

7 September 2026

Back

Has Steam brought AAA gaming to Linux?

Exploring Valve’s Steam Deck, Steam Machine, and the Future of Linux Gaming

2 July 2026

Gaming on Linux has long been the underdog story of the tech world. Windows has dominated AAA game development for decades, thanks to its tight integration with DirectX, the go-to toolkit that makes characters pop, physics run smoothly, and graphics shine. But Valve, with its Steam Empire, may be changing all that.

Steam’s bold move: Linux gaming gets a boost

Valve’s Steam OS, built on Linux, powers devices like the Steam Deck and the Steam Machine. By cutting out Windows licensing fees, Valve can focus resources on performance, giving gamers more power for the things that matter most—the games themselves. Historically, Linux gaming was tricky, like showing up with Marmite sandwiches while everyone else had cheese and ham. Now, with Valve bringing AAA titles to Linux, the game is changing.

Proton: The unsung hero

The real magic lies in Proton, Valve’s translation layer built on Wine. Proton converts Windows’ DirectX calls into Vulkan, Linux’s preferred graphics API. It’s like putting on a pair of magic glasses: most games run smoothly, but sometimes the translation stumbles. Despite occasional quirks and minor performance overheads, Proton allows popular AAA games like Cyberpunk to run on Linux devices without major headaches.

The anti-cheat hurdle

While Proton bridges the DirectX gap, anti-cheat software remains a sticking point. Programs like EasyAntiCheat and BattlEye operate in Windows’ kernel space, a restricted area Linux doesn’t readily allow. This limits Linux’s ability to run competitive multiplayer titles securely, leaving some AAA games inaccessible unless Windows is installed.

The path ahead

Valve has made major strides: standardised distribution with Flatpak, Vulkan performance gains, and a vibrant modding community. Cloud gaming also promises to lessen OS limitations, opening the door for broader Linux adoption. Still, anti-cheat challenges and developer hesitation mean the journey isn’t over.
What do you think? Is Linux finally ready for AAA gaming, or are anti-cheat and DirectX hurdles too high for one company to overcome?

Join the conversation and let us know your thoughts!

Watch our Lesson Hacker video to find out more.

For more great Lesson Hacker videos, check out the CraignDave YouTube playlist HERE.
Visit our website to explore more cutting-edge tech news in the computer science world!

Related posts

Beyond the IDE: programming lessons in the age of AI

One of the most persistent challenges in a programming classroom is knowing what every student is doing at any given […]

23 October 2026

If a NOT gate turns zero into one, where does that one come from?

How does a NOT gate turn a 0 into a 1 without creating something from nothing? Discover how voltage, transistors and digital logic work together to power the technology inside every modern computer.

15 October 2026

Ofsted after the Big Listen: have the rules changed, or just the packaging?

Ofsted has changed its inspection framework, but many school leaders question whether it has changed the culture surrounding inspection. We explore whether report cards can rebuild trust, reduce pressure and create an accountability system that genuinely helps schools improve.

9 October 2026

Should I make my own resources?

Should teachers really be making all their own resources? We explore the balance between creating, buying and adapting resources — and why thoughtful planning matters more than who made them.

5 October 2026

How does age verification technology work?

Can technology really tell how old you are? Explore the computer science behind age verification and the privacy and accuracy questions it raises.

28 September 2026

UTF-8 is witchcraft, right?

Ever wondered how your computer can happily store a simple English letter, an emoji, a pound sign and characters from […]

22 September 2026

Why teaching models keep returning

Teaching frameworks can provide valuable guidance, but when they become rigid formulas, they can undermine professional judgement. The goal isn’t identical lessons – it’s consistently effective teaching.

18 September 2026

Do GCSEs doom students to fail?

GCSE results may look stubbornly flat, but the reality behind the numbers tells a very different story. Why has further improvement become so difficult, and are we measuring success in the right way?

11 September 2026

Will Quantum Computers Break Encryption?

Could quantum computers really break encryption and put our data at risk? We explore the reality behind the hype, from qubits and Shor’s Algorithm to the future of post-quantum security.

7 September 2026