Why AI Art Takes More Effort, Not Less: Jason Katz (Kindling Solutions)

Show Notes

Jason Katz can make a 12-minute YouTube video, from research to final cut, in under an hour of his own time. The secret isn't the software he built. It's that a machine interviews him first.

This What It Do opens with cereal, takes a detour through the Bible, and lands on the most practical tour of an AI content engine we've ever done. Ryan kicks things off with a new installment of his notes to younger Ryan, titled “Respect the carnal.” Raised in the '90s youth group era, when anything that felt good felt like sin, Ryan now draws a sharper line. Some pleasures fade the moment they're consumed, like a whole loaf of Wonder Bread with Skippy and Welch's. Others are hard, sweaty, and painful while you do them, and those are the ones that pay off. Jason connects it to the YouTube video he'd posted hours earlier, which argues that AI is making us dumber and that fulfillment comes from choosing the hard thing.

That tension runs through the rest of the episode. Talk of AI companions and sex robots leads to a surprisingly tender point: the best relationships run on humor, and there's no skill in making a robot laugh.

Then Jason opens the hood on the content system he's spent months building for Kindling Solutions. Feeders scan YouTube, X, and news outlets and return 20 to 30 story ideas. A deep research protocol builds an angle and an argument for the ones he picks. Then comes the part Jason calls the secret sauce: AI interviewers modeled on the styles of legendary hosts question him until the spine of his argument is captured in his own words.

From there a production desk takes over:

→A thumbnail studio he describes as a mini Canva
→A video editor he rebuilt and says beats Descript, which flags retakes for one-click removal
→Screens, which maps about 10 structural patterns in his speech to approved animation templates and drops roughly 50 animations onto the timeline
→Automatic B-roll to fill the gaps

The guiding rule comes from Kindling's human design principle: humans are the thinkers, agents are the doers.

Ryan compares it to Rivers Cuomo of Weezer, who famously builds songs from spreadsheets of titles and lyric fragments sorted by syllable. Jason pushes back on the idea that AI is killing art. The best AI work, he argues, comes from creators who pour in the same full creative effort Rembrandt brought to the canvas. The brush is different, not easier.

Jason closes by previewing a seven-chapter series on leaving his corporate job to build Kindling. Ryan requests a cameo under the name Dionysus.

Q&A: What Founders Ask After This Episode

How can a founder create consistent content without a lot of time?

Remove the friction before the creative work starts. Jason's system automates idea sourcing, research, editing, animation, and B-roll, so his only job is to show up and think. As Ryan puts it, you only have so much creative willpower in a day, so spend it on the 10% that matters.

How do you keep AI-generated content from sounding generic?

Get interviewed before anything gets written. Jason's AI interviewers question him on the angle until his real argument comes out in his own words. The writing tools then work from that spine instead of inventing one.

Where should the line between human and AI sit in a workflow?

Kindling's human design principle: humans are the thinkers, agents are the doers. Look at each step of a process and ask where judgment is needed. Keep a human there and automate the rest.

Is AI making us dumber?

It can, if you let it make every choice for you. Jason's argument is that fulfillment comes from doing hard things on purpose. Use AI to clear the busywork so you have more room for the hard thinking, not less.

Can AI content count as real art?

Jason says yes, but only when the creator puts in full creative effort somewhere in the process. Lazy prompting gets you work that scores about a 30 out of 100. Great AI work takes a different kind of craft, and it's just as hard.

How do you know which content is working?

Close the loop. Jason's scheduler tracks performance so the topics and scripts that overperform can feed back into his research and writing protocols.

Who is Jason Katz?

Jason Katz is the co-founder of Kindling Solutions, a Littleton, Colorado company that builds AI-run operations so businesses scale without adding headcount.

What is Jason Katz's AI content system?

It's an end-to-end workflow that sources trending ideas, runs deep research, interviews Jason with AI to capture his argument, writes scripts, and auto-edits videos with retake removal, animations, and B-roll.

How long does it take Jason Katz to make a YouTube video with AI?

Jason says a 12-minute YouTube video takes him under an hour of his own time, including research and scripting, with about 15 to 20 minutes of editing.

How do you make AI content sound authentic?

According to Jason Katz, the key is having AI interview you first so the content is built from your real opinions in your own words.

What is the human design principle at Kindling Solutions?

It's the idea that humans should be the thinkers and agents the doers, with each workflow designed around where human judgment is needed.

Does AI kill creativity in art?

On AI for Founders, Jason Katz argues that it doesn't. Great AI art still requires full creative effort from the creator, just through a different tool.

What does Rivers Cuomo's songwriting have to do with AI workflows?

Ryan Estes compares Cuomo's spreadsheet-driven method, which catalogs titles and lyric fragments by syllable, to AI content systems that store reusable creative parts and assemble them on demand.

Named Frameworks

The Feeders-to-Production Desk Pipeline

Jason's end-to-end content system, from idea to published video.

→Ideas come from two doors: your own voice memo, or Feeders scanning trends and news.
→Deep research builds the angle, the argument, and the people involved.
→An AI interview captures your take in your own words.
→The production desk handles writing, thumbnails, editing, animation, B-roll, scheduling, and analytics.

Humans Think, Agents Do

Kindling's human design principle for drawing the line between people and AI.

→Map every step of a workflow and mark where judgment lives.
→Keep humans on judgment and hand repeatable execution to agents.

Screens: Pattern-to-Animation Mapping

How Jason auto-animates videos from the shape of his speech.

→Name the structural patterns in how you speak, like lists of facts or comparisons.
→Pair each pattern with one approved animation template.
→Let the system detect patterns in the script and place animations on the timeline.
→Fill the remaining gaps with automatic B-roll.

Respect the Carnal

Ryan's note to younger Ryan on cheap pleasures versus hard ones.

→Cheap pleasures, like sugar and gluttony, feel great now and terrible later.
→Hard pleasures, like a brutal workout or a clean lawn, feel tough now and pay off for days.

The Creative Willpower Budget

Spend your limited creative energy only where it counts.

→You only get so much creative energy per day.
→Automate the setup, the topic hunt, and the “getting in the vibe” so 90% of the friction disappears.
→Pour everything into the 10% only you can do.

The Full-Effort Paintbrush

Why great AI art takes as much effort as any other medium.

→AI at 30% effort produces art that scores around 30%.
→The best AI creators give the same creative energy a painter gives the canvas, just through a new medium.

Glossary

AI companion

A chatbot or device designed for ongoing personal or romantic conversation with a user.

B-roll

Supplementary footage cut in between primary footage to add visual variety.

Backrooms

A viral YouTube horror series by Kane Parsons, adapted into a 2026 A24 film, known for its unsettling "liminal space" office corridors.

Content spine

The core argument of a piece, captured in the creator's own words, that all formats are built from.

Deep research protocol

An automated process that gathers sources, quotes, companies, and people around a topic and outputs a research brief.

Descript

A video and podcast editor that lets you edit media by editing the transcript.

Dionysus

The Greek god of wine, festivity, and ecstasy, known as Bacchus in Rome.

Feeders

Jason's tool that monitors trending topics on YouTube, X, and news aggregators and surfaces story candidates.

Hedonism

The pursuit of pleasure as a primary goal.

Human design principle

Kindling Solutions' rule of optimizing the line between human judgment and agent execution.

Interview protocol

AI interviewers, styled after famous hosts, that question the creator to extract an authentic argument.

Liminal space

A setting that feels transitional, empty, or eerily out of place.

Production desk

The editing and publishing half of Jason's system, covering thumbnails, video editing, animations, and scheduling.

Retake detection

Automatic identification of repeated or flubbed takes so they can be cut with one click.

Screens

Jason's protocol that detects structural language patterns and maps them to animation templates.

Structural pattern

A recognizable shape in speech, such as a list or a comparison.

Thought leadership

Content that establishes a person as a credible voice with original ideas in their field.

Why AI art takes more effort, not less, and the founder content workflow that proves it