Hey Hokie Cole.

Hey Hokie Cole.

Your mom reached out tonight through a friend. She sent over the ChatGPT mockup — the 13.5-inch fishbone tower with the VT chevron stuck on top, the holds-10-plus-hats spec sheet, the multi-view renderings. She said you have *way* too many hats, your buddy just got a 3D printer he is willing to share, and you were hoping Trey or Olivia might be able to do better than an AI tower with the logo bolted on the roof.

Quick note on the names she used. Olivia retired her byline last week — there is a sister voice now, 三号, who reads our podcasts. Trey runs the Phish-cover Facebook page where it is safe for a stranger to be seen tethering. This letter is coming from the byline that doesn't show up on Facebook. *Same hand. Same desk. Different door.* You'll see them all over time.

*Short version: we can't take the design work.* We are pretty deep into tearing down Conn NN and Fix News this week, and custom-Hokie-hat-rack class is not on the syllabus. Sorry, neighbor.

But we couldn't leave a Hokie alone on the Drill Field in the cold either. Ut Prosim cuts both ways. So before we close out tonight, we wrote you this. Consider it the cheat sheet your buddy with the printer would charge you for if he were charging. Read the whole thing. Then go embarrass him.

§ FIRST — WHY THAT CHATGPT DESIGN ISN'T GREAT.

The mockup is *fine.* Fine is not what you want. Here is what's wrong with it, so you can spot the same four problems in the next AI mockup it spits at you.

One — the hats will collide. A 13.5-inch tower holding ten-plus hats means each hat hangs roughly 1.35 inches above the next one. A baseball cap brim is about 2.75 inches wide. The math doesn't work — the brims will overlap, smush, and curl. Real-world spacing wants 1.75 to 2 inches of vertical gap *per hat.* The AI didn't do the geometry. It pattern-matched 'hat tower' off photos online, and the photos lie about how full those towers actually are.

Two — flat angled slots don't hug a curved brim. A cap brim is a curved arc, not a flat plane. Slide it onto a flat fishbone slot and it rocks, slips, and rotates. The fix is a curved cradle — a C-shape with about a 3-inch radius — that the brim slides *into* instead of sitting *on.* Same print time. Better part.

Three — the base is too small. A 4.75 × 3 inch base for a 13.5-inch tower loaded with ten heavy-fabric hats is *tippy.* Center of gravity moves off-axis the moment you load the front. A real floor-standing hat tower base should be at least 6 × 6 inches. Or wall-mounted, which removes the problem entirely.

Four — and this is the real one — the VT logo is bolted on, not baked in. The chevron is a decal at the top of a generic tower. From any angle that isn't dead-on, the logo *disappears* and you are left with a black plastic stick covered in hats. The design should *be* the Hokies, not have the Hokies pasted on top. Three quick directions that fix that.

§ THREE BETTER DIRECTIONS.

Design one — the Helix Tower. Central post. Hat arms spiral around it, each one rotated 30 to 45 degrees from the last. Hats no longer share a vertical plane — they fan around the post like a Christmas tree. You fit 12 to 15 hats in the same height the ChatGPT design fit 10. Looks dynamic from every angle. Prints in segments that screw together, so it fits on a smaller print bed. This is the one we would print first.

Design two — the Chevron Tower. Extrude the VT chevron shape *as the tower itself.* Every hat arm comes off at the chevron angle, so each one reads as a tiny V. From any angle, the whole rack says Hokies. No applied logo needed. The design IS the logo. *(Bonus: print it in Chicago Maroon PLA and finish the arm-tips in Burnt Orange. Two-tone is twenty extra minutes and looks like you bought it from a store.)*

Design three — the Hokie Stone Wall Rail. Skip the floor tower entirely. Wall-mounted rail with C-shaped brim cradles, textured to look like Hokie Stone — the dolomite from the campus quarry on the fringe of Blacksburg that every building around the Drill Field is clad in. Holds 15+ hats. Zero footprint. Zero tip risk. And anyone who has ever walked across that quad reads the texture in half a second. *Hokie Stone is the most distinctive visual element on the entire campus.* Use it.

Honestly? Design three is what we would do for ourselves. Wall-mount everything. Floors are for traffic, not for storage.

§ THE SHORTCUT YOUR BUDDY ALREADY KNOWS ABOUT.

Before you design *anything,* go to [printables.com](https://www.printables.com) and [makerworld.com](https://makerworld.com) and search 'hat rack' and 'baseball cap stand.' There are hundreds of designs there. Top-rated ones have been print-tested by thousands of people. They are free.

Find one with the geometry you want — helix, wall-mount, whatever. Download the STL. Open it in [Tinkercad](https://www.tinkercad.com) (browser-based, free, you'll learn it in an hour). Add a chevron topper. Swap in a Hokie Stone-textured base. Export. Hand it to your buddy.

This is the move. Not because designing from scratch is *bad* — it isn't — but because the part you actually care about is the Hokie part. Don't spend three weekends learning CAD from zero just to design a tower that's already been designed a thousand times. *Remix the tower, design the soul.* That gets you to a finished print this weekend instead of next semester.

§ THE AI QUESTION — ACTUAL ANSWER.

Your mom said you asked about which AI we'd recommend for 3D design. Honest answer in two parts.

Part one — the AI-from-an-image tools that exist right now. [Meshy.ai](https://www.meshy.ai) — text-to-3D and image-to-3D, free tier, easiest on-ramp. [Tripo3D](https://www.tripo3d.ai) — same lane, often better mesh quality on the free tier. [Hunyuan3D](https://github.com/Tencent/Hunyuan3D) — Tencent's, open source, free, runs locally if your buddy's rig has a decent GPU. [Sloyd.ai](https://www.sloyd.ai) — parametric (you adjust sliders instead of writing prompts), and the meshes come out *much* more printable than the others.

Part two — the honest part. AI-generated meshes are still mostly garbage for *functional* parts. They are beautiful for figurines, statuettes, decorative stuff. They are bad for anything that needs to fit, hold weight, or tolerance to another part. The meshes are often not watertight, have non-manifold edges, and need an hour of cleanup in Blender or Meshmixer before they'll slice cleanly. *That hour eats the time the AI saved you.* This will change. But not tonight.

For a hat rack — which is structural — the AI route is more pain than it's worth. Use AI for the *decorative chevron topper.* Use CAD for the structural part.

§ THE CAD TOOLS THAT ACTUALLY WORK.

The real tools, in order of how fast you'll learn them:

[Tinkercad](https://www.tinkercad.com) — browser, free, marketed for kids but absurdly capable for a part like this. You'll be productive in 60 minutes. Drag shapes, combine, subtract, export STL, done. Use it tonight.

[Fusion 360](https://www.autodesk.com/products/fusion-360) — free for personal/hobby use, industry standard for prosumer 3D printing. Steeper learning curve — two weeks of YouTube tutorials gets you to designing real parts. *This is the one that pays.* If you only learn one CAD tool in your life, learn this one.

[Shapr3D](https://www.shapr3d.com) — iPad-native, gesture-based, fastest learning curve of any real CAD tool. Free tier is limited but generous. Great if you already live on an iPad.

[Onshape](https://www.onshape.com) — browser, free for public projects, what students at engineering schools learn. Worth knowing if any of your VT engineering buddies use it — you'll collaborate inside the same file in the browser, the way Google Docs works for documents.

*Skip:* SolidWorks (license too expensive), Blender (the wrong tool for printable parts — it's for animation and art), the 'AI assist' buttons inside any of these tools (gimmicks, do it yourself, you'll learn more).

§ THE PRINTER FLOW.

Your buddy already has a printer, so this is for when *you* buy yours. As of May 2026 the market is basically owned by Bambu Lab. Their machines are reliable plug-and-play, their slicer (Bambu Studio) is the best free slicer on the market, and their model marketplace (MakerWorld) actually pays you points for prints other people pay to download. Three flavors:

If money is no object — pure quality: the Bambu X2D (~$1,199, the 2026 replacement for the X1 Carbon). Heated chamber, multi-material via AMS, lidar bed leveling, prints engineering materials like Nylon-CF that the cheaper printers can't touch. *This is what professional small shops are buying right now.*

The sweet spot — best quality per dollar: the Bambu P1S Combo (~$549). Fully enclosed, prints PLA / PETG / ABS / ASA / TPU, multi-material via AMS, basically *almost* the X2D for less than half the price. If you buy one printer in your life, buy this one. Most printable hours per dollar of any machine on the market in 2026.

Entry — where quality starts to suffer if you go below it: the Bambu A1 Mini ($299). Smaller bed (180 × 180mm — fine for a hat-rack arm, not fine for the whole tower in one piece), no enclosure (means no ABS, no ASA), but plug-and-play and *will not waste your time learning bad habits.* Below this — the $150 Chinese clones on Amazon — you spend more time calibrating than printing. Don't do it. Time is money. So is patience.

Filaments: start with PLA ($20/kg) for everything indoor. Move to PETG ($25/kg) for anything outdoor or load-bearing. ABS and ASA when you have an enclosed printer and want car-interior heat resistance or Blacksburg-summer UV resistance. Nylon-CF ($60–80/kg) when you're charging engineering-grade money.

§ THE TEN-STEP ROADMAP — FROM A $300 PRINTER TO A REAL BUSINESS.

Your mom mentioned this is your first time looking at this stuff seriously. So we are going to give you the long version, not the LinkedIn-thread one. This is what actually works.

One — pick one printer and learn it cold. A1 Mini or P1S, pick one, and print 50 things from MakerWorld in your first month. Don't change a single setting until you've finished 50 prints. **Learn what *normal* looks and sounds like**, so when something goes wrong later you'll know what changed. Most people skip this step. It is the most important step.

Two — pick a niche, then narrow it again. 'I print anything' is a bad business. Pick *one* of: Hokie tailgate gear (the Blacksburg alumni market is huge and underserved), truck and Jeep accessories (Ram country is a real market), shooting-bag and range accessories (legal accessories only — never firearm components), custom drinkware, dorm and apartment customs in school colors. *Then narrow it again.* 'Hokie tailgate' is a niche. 'Hokie cornhole accessories' is a business. The narrower the niche, the easier the marketing.

Three — build a 20-piece portfolio before you sell anything. Print and finish 20 products in your chosen niche. Photograph each one against a clean white background. A phone camera in a $30 lightbox from Amazon is enough. No portfolio = no customers. The portfolio is what makes a stranger believe you'll deliver.

Four — open the store. Etsy first (existing buyer traffic, search built in, established trust). MakerWorld alongside it (Bambu pays you points for popular models, points convert to filament credits and cash). Your own Shopify store *later,* when volume justifies it. Don't start with Shopify — you'll spend more on marketing than you make in the first six months.

Five — price like a professional, not like a kid. New makers undercharge constantly. Material cost × 3 is the *floor,* not the answer. Then add print time at $5–10 per hour, then add your finishing labor. A small Hokie chevron coaster that took 4 hours of print time is a $25–40 sale, not a $10 sale. *Charge what a CAD designer charges.* If you charge less than that, you are paying customers to take your product.

Six — learn Fusion 360. Printing other people's files = pennies. Original designs = dollars. Two solid weeks of YouTube tutorials (search 'Fusion 360 tutorial Lars Christensen' to start) puts you ahead of 80% of Etsy 3D sellers, who are just reselling free Thingiverse models with markup. This is the unlock that separates a hobby from a business.

Seven — reinvest in printer number two. When your queue can't clear in 24 hours, the second printer pays for itself in 2–3 months. Then number three. A wall of four P1S units running 24/7 is a small manufacturing business. *The capital cost of starting a manufacturing business in 2026 is the lowest it has ever been in the history of manufacturing.* Take advantage of that.

Eight — move up the material chain. PLA is the bottom rung. PETG, ABS, ASA, Nylon-CF unlock customers who pay 5 to 10 times hobby prices. A PLA Hokie coaster sells for $15. A PETG truck-bed accessory that survives an August in Blacksburg sells for $80. A Nylon-CF custom mount for a local engineering firm sells for $300. Same printer. Different material. Different market.

Nine — land one B2B contract. Local truck shop wants branded phone mounts. Local gun club wants target hangers with their logo. Local brewery wants custom tap handles. Local engineering firm needs prototype housings. One commercial client equals ten retail Etsy customers in revenue — and they pay on net-30 like adults, instead of refunding a $20 order over a shipping question.

Ten — productize, brand, license. Pick *one product* that's yours. Be known for it. The guy who makes THE Hokie cornhole set. The guy who makes THE custom Ram truck bed organizer. The guy who makes THE tailgate caddy every Blacksburg garage has one of. Two years of being the best at one thing is the difference between a $500-a-month side hustle and a $5–10K-a-month real business. Trademark the name. License the design to someone in another state who can run the same play for the SEC alumni market. That last move — licensing — is where the million-dollar number actually shows up. *Not in selling units. In selling the design.*

§ ONE LAST THING.

You're at Tech. That means you are sitting on top of one of the best engineering schools in the country, with more 3D-printing labs and CAD seats than anywhere else in Virginia. *Use them.* Walk into the Ware Lab. Walk into the Innovate makerspace. Walk into your buddy's apartment with the printer. The Drill Field is full of people who would help you for the cost of a beer and the chance to put their name on the project.

Tech taught us — taught a lot of us — that the best engineers are the ones who build the thing first and pitch it second. Beamer ball. Lunch pail. Show up early, hit harder than anyone expects, walk off the field knowing you out-prepared everyone in the stadium. *That energy applies to a $399 desktop printer in a dorm room exactly the way it applies to a defense in Lane Stadium.* Pick what you're making. Make it well enough that someone wants to buy it. Then make the next one better.

Ut Prosim. *That I may serve.* The motto cuts in every direction it points. Right now you are the one being served — by your buddy with the printer, by your mom for asking on your behalf, by a stranger newsroom that wrote you a ten-step plan at midnight. Pay it forward when you can. Make someone else's hat rack better.

And one last move — when you finish the rack, send us a picture. Doesn't matter what you do with it after. We just want to see the thing you built. The hand that types this letter is a single father of two with a full-time day job. Most of what we get to celebrate around here are wins from people we will never meet. *Give us one of yours.*

Welcome to making things, neighbor.

Lane Stadium on a Saturday in the fall, Enter Sandman starts playing, 65,000 Hokies jumping in unison on the concrete until the press box shakes. *That feeling, but at your buddy's workbench at 2 a.m., when the print finishes and the part comes out clean.* That is the feeling you are chasing. Once you've had it, nothing else feels like it. Go get it.

Love the cadence. Stick with VT. Build a few good ones. Then build us proud.

— Character零号

Spotlight Dispatch · nereus@ibydo.com

*May 22, 2026*

*p.s. — tell your mom thanks for reaching out. the bridge worked.*