Fast Hours

Opus 5.5 made editable AI video practical: build motion graphics in code, then revise shots, text, and variations without starting over.A cocktail graphic from Rory’s T-shirt became an animation system. Then came four more drinks, a Pokémon battle, podcast promos, and a client-style video. Drew and Rory sound like two people who’ve spent years hearing “one tiny change” and knowing exactly how much work that used to mean.Claude Opus 5.5 turns one cocktail image into a JavaScript or HTML animation system and produces four more recipe videos. Three.js lets Rory block a Pokémon battle and change camera angles for complex scenes. Riverside supplies context for Fast Hours promos. Drew tests a client-style video using a website and company context file.#Opus55 #AIVideo #MotionGraphics #claudeai ---⏱️ Fast Hour00:00 What are Drew and Rory making this week?07:31 Why is AI adding work for project managers?9:29 Why most companies are stuck at basic AI17:36 What is production debt, and why AI took off20:45 Is the $200 AI plan a different product?21:19 What changed with Claude Opus 5.525:04 How one T-shirt image became a motion graphic26:28 Why code-rendered video is infinitely editable28:19 How to turn one render into an animation pipeline31:25 Can Opus 5.5 render a 3D Pokémon battle?36:58 How can Three.js block a complex shot?41:36 How did AI make a Fast Hours promo?42:28 Do you need brand guidelines for this?45:41 Can AI pick a stronger clip hook?47:34 Cutting a promo with no instructions49:18 Why creatives aren't siloed anymore50:39 Watching a sponsor promo render live53:38 What one Opus 5.5 prompt gets you for a B2B SaaS video56:09 The exact prompt behind the one-shot video57:21 Claude Code vs Cowork for motion builds59:17 How do editing and motion graphics combine?1:03:05 How did Drew turn his kids' handwriting into a font?1:10:33 Stop sending AI output without the context1:14:17 How to share a skill that runs anywhere1:18:40 Why trading transcripts beats trading drafts1:20:40 Can you share a Claude project?1:24:18 How to get email discounts, then unsubscribe1:27:44 Where does the episode close?

After 20 years in creative, building Golden Wolf and selling it to Doodles, Ingi Erlingsson went deep on AI to start a new chapter at his company, SYSTMS.Ingi has spent two decades making work for brands including Nike, Adidas, Jordan, Disney, Netflix, Coca-Cola, and Cartoon Network. He built Golden Wolf from the ground up, sold the studio to Doodles, became Doodles’ Chief Content Officer, and now builds custom AI creative systems at SYSTMS.So naturally, Drew and Rory spend half the episode staring at his ComfyUI graphs wondering what the hell is happening.Ingi breaks down how he actually works with AI: custom ComfyUI nodes, LoRAs, agents, model fine-tuning, Wan 2.2, Krea 2, GPT 2.5, After Effects, transition models, and the reusable systems behind his visual work. The conversation also gets into creative control, originality, attention, open source, prompt hunting, and why experienced creatives may have an advantage as AI gets more capable.More importantly, Ingi explains why the goal matters more than the prompt, why rapid iteration changes the creative process, and why giving AI more capability does not automatically mean giving it more creative decisions.#GenerativeAI #ComfyUI #AICreative #creativedirector ---⏱️ Fast Hour00:00 Who is Ingi Erlingsson?02:58 How much creative freedom do top brands give?05:56 How do studios get hired for their style?10:59 What did Doodles teach Ingi about brand work?13:05 How do you turn attention into recognition?18:13 How does Ingi build his signature AI style?26:13 Why does AI make experimentation cheaper?29:01 How do AI and After Effects work together?30:20 Why did Ingi move into ComfyUI?31:42 What hardware does Ingi use for ComfyUI?34:34 What does ComfyUI let you control?38:33 How does Ingi create tracked AI graphics?42:05 How do you build custom ComfyUI nodes?47:11 How does Ingi combine fine-tuned AI models?51:09 How should you train a style LoRA?55:13 What should a video LoRA learn?58:50 How much creative control should you keep?1:01:04 Do AI videos need longer generations?1:07:28 How can open-source AI sustain itself?1:11:31 What did Doodles and Web3 teach him about AI?1:15:01 How does viral content connect to X payouts?1:17:42 AI critics keep moving the goalposts1:19:12 How do bots expose themselves in replies?1:21:48 How does AI search surface a local business?1:24:04 Why do comment-for-workflow posts spread?1:25:42 What should young creatives learn now?1:28:23 What should AI creatives get good at?1:32:15 Do creatives still need to collect prompts?1:32:54 How did Ingi end up working on DuckTales?1:36:12 Who gets the Fast Hours sign-off?---🔗 Links Ingi's X: https://x.com/ingi_erlingssonIngi's Instagram: https://www.instagram.com/ingi.ai/Ingi's company (SYSTMS): https://www.systms.ai/

So, what can OpenAI’s Astra and GPT Image 2.5 actually do?
Drew Brucker and Rory Flynn put OpenAI’s Astra and GPT Image 2.5 models through a golf course build, a vintage typewriter animation, and a round of image edits. Rory’s golf course experiment burns 70% of his weekly allowance. Drew spends three hours getting a typewriter to type. Their gift for turning a quick test into a side project remains undefeated.
The pro-AI half of the show. The other half: a viral AI resignation that looks planned to the minute, Anthropic's 154-page report on how people misused Claude, and GPT Image 2.5 quietly fixing the reasons everyone kept running back to Nano Banana.
This episode covers OpenAI Astra with computer use, two ChatGPT chats collaborating inside one project, Three.js and Blender renders, a fantasy football analytics hub, and Astra running Midjourney with mood boards and style references. GPT Image 2.5 multi-edit comments, sketch input, custom aspect ratios, and template creator. Anthropic's threat report, Moonshot and Kimi distillation, and the viral AI researcher resignation.
---
⏱️ Fast Hour
00:00 What’s on Fast Hours this week?
06:35 Why do traumatic moments burn into memory
07:42 What was different before social media
13:48 This week: Astra, GPT Image 2.5, a viral resignation
15:04 Did an AI researcher really resign over ASI
18:40 How to tell what is true in AI news
20:42 Why quitting your AI job can't stop superintelligence
27:01 What do the cited reports allege about Kimi
28:33 What AI misuse examples do the hosts discuss?
34:07 How did Astra build a golf course prototype?
35:35 How did research improve Astra’s first attempt?
36:33 What happened when two AI chats collaborated?
39:29 Why did Rory stop the drainage research?
40:02 How did Drew approach his Blender experiment?
40:52 How did Astra recreate Drew’s typewriter?
43:49 How long does AI-assisted 3D work take?
44:43 How did Drew build a ChatGPT dashboard?
48:32 What happened when Astra explored Midjourney?
53:56 How did mood boards shape Astra’s prompts?
58:49 What improved in GPT Image 2.5?
01:00:24 How do GPT Image 2.5 edit comments work?
01:01:47 How do you request multiple image variations?
01:03:30 How well did GPT Image 2.5 resize ads?
01:07:11 How reusable templates in GPT Image 2.5 can help brands
01:08:37 What’s coming next for Fast Hours?

To run a solo business on AI, build a system of small skills and agents that handle the admin, comms, and ops so your time goes to creating.
---
Two founders who swore AI would give them their time back open the hood on what they actually run.
Drew walks through 37 hand-built skills and a 21-layer system for writing in his own voice.
Rory runs his whole business out of Notion and Claude, down to finding an online fax machine for New York State. Neither of them can remember what half their own skills do.
That's kind of the point.
This episode covers building an AI operating system for a one-person business. Topics include Claude skills and agents, a voice system for writing, email and call ingestion, Notion as a CRM, red-teaming contracts and MSAs, predicting LinkedIn reach, Nano Banana prompt engineering, and handling taxes, invoices, and vendor forms. Fable and Astra model releases open the show.

An AI creative director combines design fundamentals, AI fluency, tool selection, and repeatable workflows to produce better creative work.
The boys over at Fast Hour invites a 24-year-old AI creative director to give away her playbook and somehow gets reminded that buying more AI subscriptions does not count as mastery.
Jamey Gannon explains why AI is a skill, what top operators do differently, how Midjourney mood boards become brand systems, and why sometimes the smartest AI move is opening Photoshop.
Jamey Gannon explains how self-taught design, social media, freelancing, and AI fluency shaped her work as an AI creative director. Jamey Gannon demonstrates Midjourney mood boards, SREFs, Midjourney V8.2 Edit, Krea K2, Claude, Figma, Photoshop, Flora, Nano Banana, and the CXO brand system, including when to build custom tools instead of forcing an existing workflow.
---
⏱️ Fast Hour
00:00 Who is Jamey Gannon?
06:43 How Jamey taught herself design at 15
12:23 Why Gen Z chases money over job titles
15:35 What first freelance clients look like at 17
18:43 Why social media is a freelance prerequisite
22:16 Freelance math: 20 clients vs one salary
24:13 How her sister makes $5K–$10K/month with AI
27:08 What is an AI creative director?
28:58 What the AI Creative Director course teaches
33:05 Is Krea catching Midjourney?
35:53 The "easy, fast, cheap" AI lie
38:43 Three moments that hooked Jamey on AI
42:23 How to build brand mood boards
46:00 How one SREF cracked the brand style
46:27 Midjourney + Figma: the pixel shader pipeline
49:06 When you should not use AI
51:53 Two ways to use Midjourney mood boards
56:44 What top AI creatives do differently
59:44 Build the tool when the workflow doesn't exist
1:11:28 Midjourney V8.2 Edit: first live tests
1:15:00 What works and what drifts in Midjourney Edit
1:20:06 Midjourney Edit vs Nano Banana: identity lock
1:21:32 Can Midjourney edit text and fonts now?
1:25:23 Can Midjourney's edit model preserve logos and apparel graphics?
1:31:05 Is Midjourney inpainting or regenerating?
1:34:33 The verdict: using Nano Banana less
1:37:34 Where Nano Banana still fits