Who Is Joseph Enochs?
Joseph Enochs is the Chief AI Officer at Enterprise Vision Technologies (EVT), a firm he joined when there were nine employees and about $30 million in revenue. Fifteen years later, EVT is approaching a billion dollars, serving Fortune enterprises across healthcare, media and entertainment, retail, and banking. Along the way, Joseph became the author of O'Reilly's Video Generation with AI, a book written to open the door for anyone who wants to understand how generative video actually works, from a free Google Colab notebook to building a safe product.
He didn't start in AI. He started in the Marine Corps, in a family where military service was expected. Applied machine learning came later, with a legacy of work that includes COVID outbreak tracking for Los Angeles County, computer vision for electric utilities (helicopter and drone footage for wildfire prevention), and now a seat on the governing board of the HDF Group, the open-source stewards of large-scale scientific data formats used by NASA and others. His second O'Reilly book, on physical AI, has already been accepted.
What ties the arcs together is a single move. Joseph builds bridges. Between the creative community and the technical community. Between enterprise security teams and frontier AI capability. Between the current wave of generative tools and the next wave of multimodal, physical-world AI. The books, the day job, and the board seat are all the same work, aimed at different audiences.
The Archetype: The Bridge-Builder
The Sage
The Sage
The Return with the Elixir
Joseph's primary archetype is The Bridge-Builder, with a strong Sage component underneath. The Sage drive, build the model, teach the framework, is visible in how he structures a conversation. Ask him about AI safety and he'll give you a decision rule (bounded loops, batches, human-in-the-loop for critical actions). Ask him about enterprise risk and he'll name specific standards (ISO 42001, NIST AI RMF). Ask him about career advice and he'll cite the lineage: W. Edwards Deming to Russ Ackoff to Eli Goldratt to Gene Kim, systems theory as a working tool, not a trivia item.
But the Bridge-Builder is the primary drive, and he says it out loud: "I always wanna try to build a bridge versus try to build a wall." The book came from a conversation with his own kids, who are creatives and did not want him to write about video AI. He wrote it anyway, as a bridge: an honest technical treatment that takes the creative community's agency concerns as seriously as the diffusion-transformer math.
That's the Bridge-Builder in action. The whole book is engineered to let someone with zero budget start on a free Colab notebook, then walk them through ethics and provenance in chapter two, before any of the deeper architecture lands. The on-ramp is the ethics. That's not accidental.
"I always wanna try to build a bridge versus try to build a wall."
The Hero Match
Chiron
Chiron was the wise centaur of Greek myth, the teacher of Achilles, Jason, Asclepius, and Heracles. He lived at the edge of the civilized world, half of one species and half of another, and he used that in-between position to train the next generation of heroes. Unlike the other centaurs, Chiron was gentle and learned. He didn't lead armies. He taught the people who would.
Joseph maps to Chiron on three specific counts. First, the in-between position. Joseph sits between the applied-ML research world and the Fortune-100 enterprise world, between the creative community and the technical one, between frontier AI capability and governing-board risk management. He uses that position to translate, which is Chiron's whole function. Second, the teaching posture. O'Reilly's Video Generation with AI is explicitly a book for beginners, structured as a hero's journey (his own term for the book's arc). The teacher is building up heroes who will do the work. Third, Chiron's lineage was medical and preparatory: he taught Asclepius, who became the god of healing, and taught warriors how to survive. Joseph's second book is on physical AI, with explicit focus on drug discovery, protein folding, clinical trials, and digital twins of biological systems. The pattern repeats across centuries.
Uncle Iroh
For a modern hero match, Joseph is Uncle Iroh from Avatar: The Last Airbender. The parallels run deeper than the surface. Iroh was a decorated military general before he became a tea-brewing mentor, and he carried both identities into his wisdom. Joseph was a Marine before he was a Chief AI Officer. Iroh taught through parables and specific stories, often about people the student had never heard of (Ed Catmull selling popcorn at Disneyland is a very Iroh move). Iroh's famous instinct was to build bridges across the four nations rather than conquer them, which is exactly the posture Joseph takes between creative and technical communities.
Iroh also had a specific gift for seeing young people's potential before they could see it themselves. Joseph's advice to anxious creatives, "we need creative people, they have no fear, there should be no fear," is pure Iroh. The combination of military discipline, deep technical capability, warmth about the next generation, and insistence on ethics over winning is a rare archetype, and it reads as unmistakably Iroh.
"We need creative people. They have no fear. There should be no fear."
The Story Behind the Book (Video Generation with AI)
Joseph wrote a book about large language models in early 2023. Then a lot of books about large language models came out. He started asking what would be next.
The answer came from his kids. They are creatives. Joseph had been doing image generation on the side and vision work at the day job, drones and cameras and blur detection for wildfire prevention at the electric utility. He asked his children what they thought about AI and video generation.
They told him not to write the book.
He wrote it anyway, but the reason he wrote it changed. The original manuscript was called Making Machines Dream, and it was really a book about the place of creative people in the age of AI, with the technical material arranged in service of that argument. He went deep into the history of Disney, the stories of John Lasseter working the Jungle Cruise ride and Ed Catmull selling popcorn and balloons, because he wanted the book to tell young creative people that this moment is a door opening, not a door closing. Then O'Reilly got the manuscript and said, we think there's a technical book in there somewhere. They broke it out.
What stayed from the original vision is the hero's journey structure. Chapter one gets you excited: free Colab notebooks, small open models, prompts, settings, outputs. Chapter two steps back to ethics, data provenance, the question of what model you should actually train on. Then diffusion transformers, built from scratch. Then vision language models. Then audio and video. Chapter eight is about building a safe product around all of it, with an explicit treatment of ComfyUI, LTX Desktop, and Invoke AI (now Adobe) as reference implementations.
The provenance thread runs through the whole book. Joseph works closely with LTX, an open-source model with strict standards on where its training data comes from. For anyone who wants to make video generation a profession, not a hobby, the chain-of-rights question around training data and keyframe images is the thing to pay attention to. The book treats that question as a first-order engineering concern, not a legal appendix.
The second book, on physical AI, has already been accepted by O'Reilly. The thesis: video and audio were the first steps out of grammar and into multimodality, and the next steps are tokenizing proteins, chemistry, toxicity, manufacturability, scene boxes, open world models, and eventually digital twins of biological systems. Joseph sees this as the arc that lets AI do real good in drug discovery, clinical trials, and radiology.
The Author's Journey Meets the Technology's Journey
Marine Corps to applied machine learning to COVID tracking for LA County to electric utility computer vision to Chief AI Officer at EVT to first book on LLMs to Video Generation with AI to second book on physical AI, with a seat on the HDF Group board along the way.
Still images to video generation to vision language models to multimodal audio-plus-video to open world models and Gaussian splatting to tokenized chemistry and biology to digital twins of physical environments. The book maps directly onto this arc.
The two journeys are the same journey. Joseph writes the book from inside the transition, which is why the technical treatment lands with authority but the hero's journey framing lands with warmth.
Founder Superpowers
Translating between creative, technical, and enterprise worlds
Joseph's working life sits at a three-way intersection: he reports to Fortune enterprises on AI risk, he writes O'Reilly books that teach entry-level developers from a free Colab notebook, and he raised a family of creatives. The three audiences do not usually trust each other. Joseph has spent fifteen years learning the specific vocabulary each one uses and the specific trust-breakers each one watches for. The book is the artifact. It teaches the technical material, respects the creative concern about agency, and treats data provenance as a first-order architecture problem the way an enterprise would.
Thinking in systems, not components
The lineage he cites is specific: W. Edwards Deming, Russ Ackoff, Eli Goldratt, Gene Kim. The thesis is Ackoff's systems theory. Before you solve a problem, find what it is a part of. Then find what that larger system is a part of. Only after you have the full end-to-end scope of the solution space do you drill down and become analytical. The failure mode is picking the first problem you see and optimizing it in isolation while the real pain is in the handoff three steps upstream. Joseph uses this framework at the board level, in the book's chapter structure, and in his startup advice.
Teaching the next generation on purpose
Joseph wrote Video Generation with AI at an entry level on purpose. The second book is going the same way. He coaches founders at EVT and advises builders at the HDF Group board on the same frame: build within the standards, protect the organization, pass the knowledge forward. The pattern repeats in his career advice to his children and his audience. Follow your passion. Find your community. Find a need in your community and fill it. The CalAI founders, who were teenagers, are his go-to example. The route doesn't require a Series A, it requires taste and persistence.
What It's Like to Work with Joseph
Joseph is calm and specific. He thinks in frameworks before he reaches for stories, and when he reaches for a story, it tends to be one you haven't heard. (Ed Catmull selling popcorn at Disneyland. The Opus patch-management company acquired in ten weeks. The LTX provenance architecture.) He credits colleagues by name. He names his collaborators when they're relevant and doesn't name them when they aren't.
He is direct about risk. He will tell you when a technology trend concerns him, and he will tell you why with specific numbers (the 43-day mean time to remediate, the automated exploit loop operating on days). He is also direct about optimism. He has a tremendous amount of respect for creative people, and he says so without hedging. When the conversation drifted toward enterprise strategy, he answered the strategy question and then pulled the conversation back to the first-order concern: are we building bridges or walls.
He runs on long time horizons. The second book was accepted before the first one hit the shelves. The digital-twin thesis for drug discovery, he treats as a decade problem, not a quarter problem. He is explicit that the frontier is compressing the time between idea and moat, so founders need to move faster now than they used to, but his own planning window is still measured in years.
The lab he describes offhandedly ("I have over a terabyte of VRAM here in my laboratory") is a working engineering environment, not a showcase. If you brought him a video generation product you were building, he would ask what model you chose, where its training data came from, what your consistency strategy is across frame sequences, and how you plan to pass the enterprise safety bar. In that order.
Why This Matters (For You)
If You Want to Learn Video Generation from Scratch
Joseph wrote Video Generation with AI for you. The book assumes nothing but internet access. Chapter one has you on a free Google Colab notebook, prompting a small open model and watching it generate video. The quality will not be the state of the art, but the point is to see the mechanics. Chapter two gets you thinking about data provenance before you've invested in a stack. By the time you reach diffusion transformers, you've already built the muscles to evaluate what you're learning. If you've been intimidated by generative video, this is the on-ramp. Start with the book, run the notebooks, follow the hero's journey the chapter structure lays out.
If You're an Enterprise Architect Thinking About AI Agents
Joseph's bounded autonomy framework is the one to steal. Reversibility as the decision rule for delegation. Batches with validation for non-critical workflows. Human in the loop for anything that touches audit, separation of duties, or regulatory controls. The specific standards he aligns to are ISO 42001 and the NIST AI Risk Management Framework. Build toward those and you'll be able to sell into regulated industries without rewriting your product. The compressed timeline is the real risk: the exploit loop runs in days, the defense loop runs in weeks, and that gap closes with AI-defense tooling or it doesn't close.
If You're Early in Your Career
Joseph's advice has three parts, and he has used it with his own children and his audience. Follow your passion first, because every path is hard and only passion gets you through the hard parts. Find your community second, whether that's a video game, a sport, a craft, or a technical discipline. Then find a need inside that community and help fill it. The CalAI founders were teenagers. You don't need a million-dollar-a-month app. Ten thousand to thirty thousand people supporting you on your journey is a career. The route is slow, the compounding is real, and the first step is community membership, not credentials.
If You're a Creative Worried About AI
Joseph wrote the original manuscript for you. His kids are creatives, and the first version of the book, Making Machines Dream, was aimed at the question of whether creative people have a place in this moment. His answer is: yes, and more than ever. The technology is a digital paintbrush, not a replacement for the painter. The friction in creative workflows (fifteen tools per movie) is where AI can help you stay in the creative zone. His one request: don't dig your heels in against the technology, because that is the worst possible move. Learn enough of the stack to decide where it serves you and where it doesn't, then decide on your own terms.
Go Deeper
The full conversation with Joseph Enochs is on its way. Check out other episodes in the meantime.