theBrianSykes

Today it is Adobe Firefly... for the images + the video. Inspiration?→?can come from anywhere. This time, while watching the Netflix series: Ripley. I wanted to see how Adobe Firefly could handle creating cinematic visuals + animating them. Images ::: I am following the 12 Essential Prompt Elements structure as laid out in my book - "Midjourney: Back to Basics ? Beginner's Guide." (Not just for Midjourney folks.) Firefly Image prompt →?Cinematic film still of a charming 1960s café scene in Atrani, bustling with warmth and life. A barista serves coffee to an elderly couple seated by the window. The couple chats, sipping coffee and glancing out at the street. Cozy café with wooden tables and a view of Atrani’s narrow streets. Soft morning light filters through the windows, casting a gentle glow. Centered on the couple, with the window framing the background. Photorealistic, capturing fine details and a sense of timelessness. Nostalgic, inviting, and intimate. Eye-level, slightly to the side to capture expressions and ambiance. Early 1960s, timeless and serene. Peaceful, reflective, and sacred. Portrait, drawing focus toward the altar. Variations to convey the different scenes - was employed for each render, but I followed the same general structure as the prompt shared. For the Video generation, I followed the same general structure as recommended for Runway: Camera movement + Establishing scene + Additional details Villa Scene →? Static wide shot. The couple sits at a marble table near open doors, overlooking the Mediterranean. Sunlight filters through sheer curtains, creating a warm, dappled glow across the table and their relaxed expressions. With these, I am working with less complex subject matter (and motion) than engaged with the prior video of Kung Fu Fighting. Music created in Udio. Brian Sykes AI Lab

Minimax was revealed creatively in Rory Flynn's Kung Fu fighting post. Comical and yet charmingly →?fun. I wanted to see how Adobe's Firefly Video (beta) could handle the same challenge. It's not perfect, but it's promising. Midjourney Prompt: ? DVD still from 1980s dark fantasy adventure ninja movie, epic fight scene between a {anthropomorphic dragon, mini mech warrior} and a Shaolin Monk warrior --ar 16:9 --style raw --stylize 300 --v 6.1 (Great starter prompt Rory Flynn!) I added the 'between a {anthropomorphic dragon, mini mech warrior} and a Shaolin Monk warrior' to modify the fighter focus. Firefly Prompt: ? Epic scene, the camera follows as two ninjas punch and kick, fast intense, epic action as the two battle, cinematic brilliance Note the difference from what Rory requested from Minimax: ? epic violent fight scene, the camera follows as two ninjas punch and kick, fast violent action as the two battle to the death, cinematic brilliance Adobe flagged my render request - telling me I had to modify it, without telling me WHAT needed to be modified. A guess was the terminology, so I took out the phrase 'to the death'... still no render. Then removed both instances of 'violent' - and the renders began to work. As with all AI tools - as we explore them, we uncover inherent STRENGTHS and WEAKNESSES. We also have to remember that this is STILL early in the game for all of the AI to Video platforms - most particularly, Adobe. Is it perfect? No. Can it do some pretty amazing Image-to-Video? Yeah - you can see that in the FIRST exploration I did with the platform of the desert scene (I will link that in the commments below in case you missed it.) I will keep exploring to uncover other strengths of Adobe Firefly Generate video (beta) - and keep you in the loop. Kudos Adobe. I applaud this early run and am excited to see it →?keep getting better. Brian Sykes →?PS... For the professional Video Editors and VFX artists out there... I know what you are thinking - 'This is not pro-level results.' And, you are right. Not yet. But it's moving in the right direction. Remember... this is just the first step.

Apple engineers have just declared that AI’s reasoning is “flimsy” — but is it really that simple? In a study digging deep into AI’s capabilities, Apple’s engineers revealed that advanced language models, like OpenAI’s ChatGPT, Google’s Gemini, and Anthropic Claude, often struggle with consistent logical reasoning when faced with even slight modifications to standard benchmark problems. The research, conducted by six Apple engineers, claims that current AI systems don’t engage in genuine logical reasoning. Instead, they merely replicate the reasoning steps found in their training data. While this assertion highlights some of AI’s limitations, it’s important to remember that these models are designed to mirror patterns in data, not generate reasoning on their own. AI reasoning may appear flawed when out of its usual patterns, but it excels in delivering powerful insights when guided properly. Ultimately, understanding and refining these limitations is crucial for future AI development. Brian Sykes Want to dive deeper into this controversial claim and see the details for yourself? Check out the full article here: https://lnkd.in/e8JE3Tph #Apple #AI #ChatGPT #Claude #Gemini #reasoning

AMP creative output + RAMP learning, using AI while embracing the Human Element. As a creative professional, you’re constantly seeking ways to be faster, smarter, and more innovative. That’s where AI steps in. In my latest YouTube video, I dive into how you can use AI to amplify your skills (AMP) and accelerate your learning (RAMP). Whether you’re just getting started with genAI or looking to optimize your workflow, this video covers practical tools and techniques that will transform the way you LEARN! Learn how to streamline your process while maintaining the human element. See how I use YouTube, AI tools like NoteGPT and ChatGPT, to save time and produce better work. AI isn’t here to replace us— it’s here to supercharge what we do best. Brian Sykes https://lnkd.in/exXx9fy6 In my video, I walk through how I used AI to summarize, develop action points and create hard links to core learning objectives from the awesome YouTube posted video interview by Chris Do (from TheFutur) and Brian Collins (of COLLINS). Link to that interview is in the comments. Be sure and watch it to see Brian Collins take on the change happening in Branding & Marketing! #chatGPT #Gemini #Claude #LLMs #LearningRAMP

What if... You could watch a YouTube video - and then use GenAI to create not only a full breakdown of all the key points - but derive a hard-link to WHERE in the video you can review that detail. What if... You composed a series of these hard-link files via AI with summarizing documents that you could string together - and develop a full curriculum of structured assets - with hard-links (to see the bits that are pertinent) when you want to review or dive deeper. RAMP is how you personalize learning that YOU direct. Fast-tracking information to fill the gaps on knowledge with the power of personalized AI. You can RAMP. I did. Part 1 ? https://lnkd.in/efP7mVj9 Part 2 ? https://lnkd.in/evvdyJnA This one... just might blow your mind.

AI - from Article > Podcasts > Video (3 part series - so far)... I continue from Image to Video with Animation and Lipsync today in the Substack. Tech Stack includes: Google Gemini ? NotebookLM + Midjourney + Descript + Adobe Photoshop (Beta) + Adobe Audition 2025 + Runway Gen-3 Alpha Turbo + Adobe Premiere Pro 2025 https://lnkd.in/ew6icwNz Brian Sykes

Why GenAI is the Creative Industry’s Wake-Up Call and How You Can Lead the Charge The world of creative work is evolving faster than ever. GenAI is no longer a distant concept—it’s here, and it’s transforming everything. As Creative Directors, you’re facing a double-edged sword: ? the daily demands of leading your teams - and - ? the urgent need to adapt to the rise of AI. The challenge? Integrating AI without losing the human touch that makes our work resonate. Change is hard. Resistance is real. But as leaders, it’s up to you to guide your teams through these uncharted waters—and John Kotter’s '8-Step Change Model' offers a roadmap for doing just that. Let’s dive in. Brian Sykes theBrianSykes #genAI #CreativeDirector #ArtificialIntelligence #CreativeIndustry

Think 99% of AI writing is bad? You’re absolutely right—but it's because MOST people don’t know how to use AI - correctly. AI isn’t the problem, it’s how you interact with it. Most writers miss key steps that turn mediocre outputs into something truly valuable. If you want AI to work for you, you need to: ? Prime it properly with clear examples of your style. ? Provide context, intent, and personal insights. ? Guide it like a Creative Director, refining every output until it’s aligned with your vision. When you take control, AI becomes a tool, not a shortcut. If you’re ready to integrate AI into your creative workflow while keeping the Human Element at the forefront, join us at theBrianSykes and the AI LAB. Learn the right way to use genAI and elevate your creative game. Let’s get started today. Brian Sykes #AIWriting #CreativeProfessionals #HumanElement #AI

My first book - "A.I. Explore: Prompt Fundamentals" sold over 1000 copies. All 5 of the other books produced after, sales added together, didn't compare. There is something about a Basics, Beginner's Guide, Fundamentals, etc - that moves people WAY MORE than a deeper diver after the fact. There are still plenty of people who have not yet even given AI a fair shake... so I am in the process of developing a series of Beginner's Guides on AI applications. I started with Midjourney - and am 5 lessons in (as of today). I will compile all of these lessons into 1 book to be released July 15 (or at least, that is my plan). The flow of my teaching is to help people not just have copy-paste prompts, but rather a deep dive into what needs to be considered when prompting with ANY text-to-image platform. Today's post - 5 ??Describe - just might change the way you think about prompting. I recommend reading all the chapters in sequence: 1 ??Text Prompt 2 ??Image Prompt 3 ??Parameters 4 ??Learn to See 5 ??Describe More to come... after I release: "Midjourney - back to basics ? Beginner's Guide" I will ??produce a video course on the AI Lab + ? tackle other AI applications with their own fundamental books & video courses. All subscribers to theAILab.substack.com will get the books for FREE, and a discount on the video courses. Link to today's article is in the comments. Brian Sykes Midjourney #beginnersguide #backtobasics