This week I took a sharp turn in my Frazetta-style ultralearning challenge: moving from full-body gesture drawings into the terrifying world of faces and heads.

It’s been humbling.

I started with high hopes — thinking I’d knock out a few eyes, noses, and jawlines, and be on my way to stylised glory. Instead, I’ve discovered just how complex the human face really is. And how little margin there is for error.

• 👁️ Eyes: placement, angles, lids, and asymmetry
• 👃 Nose: structure, shading, and proportion
• 👄 Lips: volume and shape without losing gesture
• 🧠 Head construction: planes, proportions, and tilt

The Frazetta aesthetic demands strong forms and exaggerated beauty — but you can’t stylise something you can’t draw realistically first. So I’ve spent hours this week grinding fundamentals.

• Still doing occasional gesture warm-ups, but 80% of my focus is now on facial structure
• Plan for Week 3: switch to hands once I’m confident with facial scaffolding
• Final Week (Week 4): shift all energy into building full-page compositions from imagination, integrating what I’ve learned

1. Finding time. Life intrudes, and drawing takes focus. I’m managing short 20–30 minute bursts most days, but not hitting the numbers I’d hoped.
2. Tutorial quality. There are endless YouTube videos, but very few that speak clearly to beginners. Pyrko has been a lifesaver. Most others assume background knowledge or go too fast.
3. Proportions. Still the hardest part. Eyes drift. Noses flatten. Faces stretch. Every small error is visible. I’m training my eye slowly, but it’s going to take hundreds more reps.

• You can’t fake facial proportions — either you get them right, or it looks uncanny.
• Stylisation starts after structure — not before.
• The eye is more sensitive to small mistakes in faces than in limbs or torsos.
• Frustration is part of the learning curve — I just keep drawing.

• Pyrko tutorials — exceptional guidance for facial structure and features
• Mmmmonexx - excellent tutorials, although in Korean, but since I'm learning Korean this helps in two different ways!
• Loomis head construction diagrams
• Daily sketch bursts using mechanical pencil on A4 paper
• Occasional overlays in GIMP to check accuracy

Week 3 is hand week — famously one of the hardest topics in figure drawing. But like faces, they’re unavoidable. If I want to channel Frazetta’s dynamic power, I need to nail expressive hands.

I’ll keep drawing — not chasing perfection, but building mass and confidence one pencil stroke at a time. I can see progress in what I'm doing, I'm not sure that 30 days is enough to even approach a decent figure drawing. But I'm learning and that is what is important to me.

I started this week with zero formal drawing experience. I’ve never studied figure drawing, gesture, rhythm, or anatomy. This is all new. But I’m embracing the War of Art principle: aim for quantity over quality. A thousand drawings in thirty days. One step at a time.

This is the first entry in my month-long ultralearning challenge: learning to draw in the spirit of Frank Frazetta — specifically full-page pencil compositions of female figures, dynamic, stylised, and raw.

The object of this month of ultra learning is just to become better at drawing. Frank Frazetta had a life time of experience and apparently started drawing at the age of two. I have no hope of catching up and doing all the stuff he did. But if I can get to the point of drawing a human figure which looks like a human figure I'll be happy.

So with any Ultralearning project the first step is to have a well defined, acheiveable goal. So after a lot of thought, this is mine.

:: Create a full-page pencil drawing of a female figure in the style of Frazetta — dynamic, stylised, and anatomical — by the end of the month.

I need to make sure to put some constraints on the learning process so I don't get "feature creep" and stay focused on the goal. So my constraints are:

• Only female forms — to stay focused and track progress in one subject
• Only pencil on paper — no colour, no digital shortcuts
• No backgrounds or fine rendering — only gesture, rhythm, anatomy, and stylised proportion
• Final output must be a coherent composition — not just sketches, but a full-page narrative figure piece

What I learned this week:

• Everything looks terrible at first. You can’t improve what you haven’t drawn. You have to draw through the bad to find the form.
• Gesture is more important than detail. Even with no anatomy, a well-placed curve suggests life.
• Rhythm matters. Lines need to dance, not just describe.
• Repetition is magic. 1000 drawings is not just a number — it’s the target that will make me draw.

I'm not going to win any awards, but I'm seeing progress from the stick figures and things in the first day.

Next week I’ll add rhythm drills (twisting poses, flowing hair, arched backs), and continue gesture studies with a tighter eye on proportion — especially thighs and shoulder width. But I really need to start to focus on faces and hands.

• Line of Action and QuickPoses.com — timed gesture drawing sites
• GIMP overlays for tracing gesture lines from Frazetta poses
• Notes from Andrew Loomis on figure construction
• Ultralearning tracking in Habitica and Org-Mode

Week 2 will be where I shift from gesture into more structured anatomy — still loose, still stylised, but with a backbone. My aim is to build up, not render down — volume through gesture, anatomy through energy.

I’ll keep breaking the pencil.

Welcome to A Year To Break My Brain (On Purpose) — my unapologetically intense journey through twelve self-designed ultralearning projects, one per month, for an entire year.

This isn’t about dabbling. It’s about deliberate, focused, high-pressure learning designed to rewire how I think, work, and solve problems — by pushing my brain to its limits.

We live in a world where information is abundant, but mastery is rare.

Like many people, I’ve spent years collecting knowledge, skills, and interests in fragmented ways — a podcast here, a tutorial there, a book skimmed but never drilled.

Ultralearning offers a different path: one that’s aggressive, intentional, and results-driven.

The term “ultralearning” comes from Scott Young’s excellent book of the same name. His work — along with others like Peak by Anders Ericsson and Make It Stick by Brown, Roediger, and McDaniel — helped shape the foundation of this blog.

Ultralearning isn’t just about learning more — it’s about learning smarter, faster, and deeper than traditional approaches allow.

Each month, I tackle a new, ambitious challenge — learning something difficult or unfamiliar with intensity and intention.

Here’s the structure I’m using for each project:

1. Define the goal clearly :: What does success look like? Whether it’s drawing 1,000 figures in a month or holding a 20-minute conversation in Korean, the target is always measurable and time-bound.
2. Research the best resources :: Before starting, I dive into books, courses, expert advice, and high-performance examples.
3. Design the learning environment :: Schedule deep work blocks, reduce friction, and optimise for consistency.
4. Drill the hard parts :: I focus on deliberate practice — isolating weaknesses and targeting them directly.
5. Test and get feedback :: I build in checkpoints, review performance, and (when possible) get outside critique.
6. Document everything :: I track effort, method, failure, and progress right here, with honesty and clarity.

These books and tools form the spine of my approach:

Ultralearning by Scott Young

The blueprint for this year.

Peak by Anders Ericsson

The science behind expert-level performance.

Make It Stick by Peter C. Brown, Henry Roediger, Mark McDaniel

How to retain and apply what you learn.

The First 20 Hours by Josh Kaufman

A practical guide to rapid skill acquisition.

Moonwalking with Einstein by Joshua Foer

Lessons in memory and mental performance.

And beyond books:

• TED Talks and podcast interviews on metacognition, accelerated learning, and flow
• Anki (spaced repetition), Habitica (gamified habit tracking), and Emacs (for building a frictionless environment)
• ChatGPT and other AI tools as research assistants, tutors, and simulated critics

Each month brings a new domain:

1. Drawing like Frank Frazetta
2. Learning to make knots
3. Make a short movie using Blender
4. Building a microservice mesh from microcontrollers
5. Learning Kubernetes, fast and deep
6. Learn to Juggle like a professional

Some projects will be practical. Some experimental. All will stretch what I think I’m capable of.

This blog is my lab notebook. It won’t always be pretty, but it will be honest.

!! This is not productivity porn. This is self-directed cognitive violence — in the service of personal growth. !!

Thanks for reading. Let’s break things — on purpose.