About
Nick is a senior technology leader with extensive experience in developing and bringing…
Activity
9K followers
-
Nicholas King shared this𝐓𝐨𝐝𝐚𝐲, 𝐁𝐞𝐡𝐚𝐯𝐢𝐨𝐫 𝐋𝐚𝐛𝐬 𝐀𝐈 𝐢𝐬 𝐥𝐢𝐯𝐞.. https://lnkd.in/gey8-4aE A Life Sciences Decision Intelligence platform designed with the purpose of enabling teams to be experts, and focusing on getting life changing treatments to the hands of health care providers, and their patients. It’s the first real expression of a bet we’ve been making for the last two years: 𝐓𝐡𝐞 𝐦𝐨𝐝𝐞𝐥 𝐢𝐬𝐧’𝐭 𝐭𝐡𝐞 𝐩𝐫𝐨𝐝𝐮𝐜𝐭. 𝐓𝐡𝐞 𝐬𝐲𝐬𝐭𝐞𝐦 𝐭𝐡𝐚𝐭 𝐝𝐫𝐢𝐯𝐞𝐬 𝐝𝐞𝐜𝐢𝐬𝐢𝐨𝐧𝐬 𝐢𝐬. Everyone’s racing to build better models. But models are already becoming commodities. The real problem is decisions. I’ve watched this for 25 years: better data better models same bad decisions Because the system those models plug into was never designed to make decisions well. So when you add AI? You don’t fix it. You just get a longer answer faster. That’s why I started Data Kinetic. And why we're launching Behavior Labs AI. Behavior Labs AI creates for each customer a secure, isolated, uniquely adapted to them, World Model that captures each step on the journey of a products lifecycle. Supporting teams of experts to be able to understand what's happening, simulate outcomes, and focus their energy at outcomes, not on updating yet another deck. Behavior Labs AI is not a model layer, it is not a wrapper. It's a system that: remembers (knowledge graph) reasons (agent mesh) simulates (synthetics) constrains itself (rules engine) proves what’s true (evidence engine) To get teams started we start with module we call Ground Truth. https://lnkd.in/gUWTiteK One molecule. One product. ~4 weeks. Everything happening around it: competitive signals regulatory movement market dynamics evidence landscape Not a report. Situational awareness that actually holds up in the meeting. Then it compounds. Modules activate. Decisions get mapped. Context stops resetting. Most companies don’t have a model problem. They have a decision architecture problem. Behavior Labs AI is live. Proudly built by Data Kinetic. Start with Ground Truth. Everything else builds from there. 𝐅𝐫𝐨𝐦 𝐦𝐨𝐝𝐞𝐥𝐬 → 𝐭𝐨 𝐬𝐲𝐬𝐭𝐞𝐦𝐬 → 𝐭𝐨 𝐝𝐞𝐜𝐢𝐬𝐢𝐨𝐧𝐬. #AI #LifeSciences #HealthTech #DecisionIntelligence #DataKinetic #BehaviorLabsAI
-
Nicholas King shared thisMichael Housman has become a great friend since I moved to Texas. He's one of the few AI experts that is able to bring together the zeitgeist of AI, and translate the cutting edge to the business masses in an understandable way. While everyone else is doing their best to make AI acceleration seem like it's one step ahead of you, he's mastered planning for the future. Big day for him with new book Future Proof launched last week. https://lnkd.in/gN8gAGV3 I recommend everyone get a copy and read this book. Gift it to you friends and family who have any questions about how AI might help their business. Also thanks to House for the few shoutouts in the book.Future Proof: Transform your Business with AI (or Get Left Behind)Future Proof: Transform your Business with AI (or Get Left Behind)
-
Nicholas King shared thisThere are two ways to look at AI in knowledge work right now. The first is the one dominating your feed: AI is coming for your job, knowledge workers are obsolete, learn prompt engineering or perish. It's loud, it gets clicks, and it's mostly wrong — or at least, it's asking the wrong question. The second is quieter but more interesting: what if AI isn't eliminating knowledge work, but forcing us to finally be honest about what counts as knowledge work and what was always just process in disguise? Think about how you actually spend your week. How much of it is genuine thinking — synthesizing information, exercising judgment, producing an outcome that wouldn't exist without your specific expertise? And how much is moving files between systems, reformatting things, sitting in alignment meetings, and navigating workflows that nobody designed on purpose? Most people I talk to put the split somewhere around 20/80. Twenty percent thinking, eighty percent operating the machine. AI is eating the eighty. And instead of mourning that, I think the opportunity is enormous. Not because "AI will handle the boring stuff" — that framing is too small. The opportunity is that for the first time in two decades, knowledge workers get to be knowledge workers again. Your value goes back to what you actually know, the judgment you've built, and your ability to look at a problem and make something happen. That's harder than following a process. It's also a lot more interesting. I wrote a longer piece on this — why the doom-and-gloom narrative is really about leverage, not about intelligence, and what it looks like when work becomes work again. Link in the comments.
-
Nicholas King shared thisWorking with Agents is an Async world. Combined with reasoning, state machines, and the willingness to delete as meetings as possible, there could be a world with just a few less excuses to catch up on email while someone reads 16pt Verdana font to the room. https://lnkd.in/gdmGBriX
-
Nicholas King shared thisLast year was a wild ride of sleepless nights and countless iterations of solving new problems, or completely rethinking old ones. Across Anthropic, OpenAI, Cursor, Google, Open Source variants, and our own SLMs my personal tokens for the year was somewhere slightly over 50b. I dropped my Cursor year in review as an image which was a nice was to see just how much we covered in a year. What's wild is that Cursor was wasn't my primary harness but it definitely was a key part of my workflow. Also their Composer model might be the unsung hero model for 2025, fast, clean, minimal exhaust, and remaining consistent over time. Today, REPL is just as much of my primary system for document and knowledge workflow as it is for development. We'll see if 2026 hits the same levels. After some prodding from folks to start sharing again, I've put together a Substack to start sharing some of what we're learning, and since LinkedIn has a character limit, it felt more appropriate to start something longer form. https://lnkd.in/g44BR2U5 Take a read. Let me know your thoughts. And onwards to 2026, the year where Applied AI value is non-negotiable, and business cases will have to be much more than just productivity metrics.
-
Nicholas King shared thisWhile everyone is losing their minds screaming the GPT5 is the end of OAI. I'm currently ranking GPT5 best in class for agentic use cases. It's so far the most predictable and hardworking of frontier class models. It just wants to build tools, which is weirding people out, but for complex use cases this behavior is proving helpful, but not for codegen which seems to be the bench most are using. Caveat, they all start off pretty strong, but then the corp overlords usually get involved and detune the models (Anthropic has the worst degradation from release of all vendors). There's actually some really interesting design choices in this, and since we rarely see a family of models come out all at once like this, this portfolio view reveals a lot about the OAI design pathway so far and what I've seen on all of their previous model releases. Looking at you 04-mini-high, which we affectionally named 'the crackie'. 𝗚𝗣𝗧-𝟱 is clearly the base model. From what I've so far, this model is the designed as both their agentic model, but most likely where the the self-reinforcement and other scaling techniques are built in. We've seen 4 scenarios where it comes up with hybrid techniques specific to a problem, rather than just a text book answer most would optimize for. This is the master training base that informs all of the others. The rise of codegen probably skewed them a little too much in performance bias this round, since GPT-5 just wants to build tools. The router I think was an attempt to be able to optimize away from GPT-5 but this failed them on launch. This is also what is weirding out folks treating ChatGPT as a Google search box. That and the session in chat seems to create a new session context each time, which makes the long chain of random prompting most folks do single context seem less intuitive - this was GPT-4o's ChatGPT masterpiece. And they broke it. 𝗚𝗣𝗧-𝟱 𝗺𝗶𝗻𝗶 is the refined, shorter available run times, likely load adjusted token limits. But also would be a good place to test incremental releases to see their impact on adaptations for future models. 𝗚𝗣𝗧-𝟱 𝗰𝗵𝗮𝘁 is the consumer focused, hide all the workings. Make cat videos, and generally be the normie safe, don't get weird with the user, model. 𝗚𝗣𝗧-𝟱 𝗻𝗮𝗻𝗼 is the efficiency, cost management model. I'm guessing this is what fuels the free tier, and probably aspires to be mobile, disconnected friendly one day. I haven't validated this with anyone at OAI, but after seeing enough portfolio views over the years, this would be my guess.
-
Nicholas King shared thisComing out of my LinkedIn Hiatus, I'm excited to share that I’ll be a joining the team at NYU to be a PANEL MEMBER at the AI Masterclass from April 25-27, 2025—a live workshop designed for Board Members, C-Suite Executives, & Senior Leaders. This masterclass brings together top AI minds, featuring Aditya (Ed) Watal — Lead Faculty, Founder & Principal at Intellibus, and 🦾Jepson Taylor — Co-Faculty, Former Chief AI Strategist at Dataiku & AI Startup Founder. This is one of my favorite forums where leaders get together and share fresh insights, connecting with like-minded leaders, and exploring how we can not just adapt to AI—but lead with it. Learn more: www.aimasterclass.com #AIMasterclass #ArtificialIntelligence #Leadership #AIInnovation #FutureReady #TechLeadershipAI Master Class for Senior Leaders, C-Level Executives and Board MembersAI Master Class for Senior Leaders, C-Level Executives and Board Members
-
Nicholas King posted thisIt's been a while since I posted on LinkedIn. Mostly there weren't enough hours in the day between building, learning, and creating new things. No predictions from me this year. But I will lay this up. A huge thanks to everyone who took the time to help us and the team last year. It means more to me every year about who takes the time to reach out, provide guidance, ask deep questions, and push the collective to be smarter. I'll always do my best to try do the same. Just to list a few of the people on my list in no particular order. But I appreciate you all. Chas Dabhia Reena Sooch Thomas W. Dinsmore Nancy Dale Stuart Jacobson Thomas Lynch Dorian Smiley Stephen Coyle Brendon Ford Taylor Campbell Alyssa Felsch 🦾Jepson Taylor Michael Housman Ragen Doyle Benavides Akash Khokha Mike Hulme Karin Jenson Zain Khalpey, MD, PhD, FACS Erik Rasmussen John Cooper and many many more who aren't on LinkedIn.
Nicholas King commented on a post
1mo
What are run!! Congrats
Nicholas King commented on a post
3mo
https://open.substack.com/pub/theaiskeptic/p/knowledge-work-is-work-again
Nicholas King replied to a comment
4mo
Zack Kielich We should all be AI skeptical, not in a way that says it's not going to ever work, or it will never beat a human at said task. But I personally don't think an omnipresent model is going to create competitive advantage alone long term. We should challenge ourselves to find ways to augment the boring things away to create more capacity for focusing on building advantages.
You raise a really interesting point on large orgs. Something I haven't shared, but I think I'll share some of it in my next post, is that over a period of time, we were able to fill in the gaps without requiring everything to be captured. In fact surprisingly I think when we added an agent specific to try and identify gaps was probably when this approach started to really make jumps. The best way I can explain it is if x + 2 = 10, x=8 or x = "eight" (the word), x = VIII (Roman numerals), x = 🎱 (the 8-ball), x = ∞ (if 2 is infinitesimally small), or x = "the number of legs on a spider, and then a human over the loop is able to validate it.
And yes last year I felt that LinkedIn had lost it's mojo and decided to just focus on research. This year, we've got a lot to share, so I decided that the longer form substack is a better way to share.
-
Nicholas King liked thisNicholas King liked thisI achieved my Master’s in Technological Futures. 🎓 academyEX 18 months ago, I enrolled just as AI was exploding into the mainstream. I had no idea how much that timing would shape the experience. The programme pushed me deep into emerging technologies, AI ethics, human-centred design, and what it means to build technology responsibly while the field is shifting in real time. For my research, I focused on something close to home: why neurodivergent secondary students can struggle to begin school assignments, and how AI might help. That became Spark, a prototype AI-tool designed around the question: “Where do I start?” A special part of this chapter has been studying alongside my son Joshua, who is on his own university journey. There’s been something pretty cool about swapping deadline stress, study hacks, and general student chaos at the same time. Grateful for what this experience has taught me, and quietly excited for what comes next. To my advisor Jade, thank you. To my cohort - what a group to have done this with 🙌 Lian Passmore Lee Palamo Chris Worth Kelly Rummins Nadine Young Brielle Raven Shourjo D. Pauline Standen Edson Koji Nishikito Patrick Marcelino Dinithi Nimasha Carol Thomson #MastersComplete #TechnologicalFutures #AIEthics #EmergingTech #EdTech *The image is AI-generated, based on me and the project. Fitting, but also slightly ick. 😂
-
Nicholas King liked thisNicholas King liked thisLet that sink in for a second. Because most organizations I work with feel the opposite. ↳ They feel overwhelmed. ↳ They feel behind. ↳ They feel like things are already moving too fast. But here’s what 17+ years of building AI systems has taught me: This is the calm part. → The models are getting better. → The tools are getting easier. → The cost is dropping. And we’re still early. I’ve sat in rooms with Fortune 500 teams who already have access to powerful AI… …and still aren’t using it in any meaningful way. Not because the technology isn’t ready. But because the organization isn’t. ✖️ The workflows haven’t changed. ✖️ The incentives haven’t changed. ✖️ The way people think about their work hasn’t changed. So what happens next is predictable. The technology will keep accelerating. And the gap between companies who adapt… and companies who hesitate… will get very wide, very quickly. That’s the part most people are underestimating. Because the advantage isn’t AI itself. It’s what happens when a team actually learns how to use it. And right now? That’s still rare. Most leaders I talk to think they have an AI strategy. Very few have real adoption. Where does your organization actually fall?
-
Nicholas King liked thisNicholas King liked thisWhat a great conversation with Nate and Andrew on The Defense Tech Underground Podcast! We covered my (winding) path from law and US Congress to Palantir to Valinor, how to source demand signal, and our new efforts in military construction. We also touched on why GTM is often neglected in Defense Tech and my best piece of advice to fellow founders: stay in the arena. Check out the full episode ⬇️
-
Nicholas King liked thisNicholas King liked thisWe built a 10-person agency that operates like 40. It was our key to world domination. We had zero intention of sharing it with anyone. But I kept giving sneak peeks to agency founders. They convinced us. Six months ago, I brought in Bert Nieves to help build AI infrastructure for a super secret Gen AI feature film. Bert left a safe exec career at Dell to start his own Gen AI company. The chemistry was instant. Explosive. We started prototyping furiously. Codename: Nexus was born. We talked to VCs. Our song & dance was OK. But it felt disconnected from where the real value was. The Goliaths and the megacorps want to get rid of all of us. The scrappy artists are suffering. Gen AI came first for the poets and the painters of pretty pictures. A tragedy unfolding at AI speed. Holding companies have unlimited budgets, unlimited headcount, unlimited compute. AI is engineered to widen that gap. The narrative says David loses this round. Then we found our David. Literally. David Gonzo Gonzalez is dangerously smart and refuses to accept the way things are at face value. Built and sold an AI company. Co-authoring the book on separating AI hype from real value. Gonzo took one look at what we'd built inside Deep Vibe and said we were sitting on something bigger than our agency. Once Bert and Gonzo started vibing with what all of Deep Vibe had been through, they saw something underneath it. The culture shift. The institutional knowledge becoming something alive. Every project feeding the next one. Not a product. A pattern. Something that compounds. Agency founders kept reaching out. Not about our creative work. About how we were working. Every conversation sharpened the same conviction. We had a choice. Go dark. Spend six months and a lot of money building alone. Risk getting disrupted by deeper pockets on Day 1. Or start unlocking real value for agencies ready to invent the future with us. Learn together. Let the people who use this shape what gets built. We chose the second door. But that only works if you can answer the harder question. Where does human value actually live when the machines keep getting faster? In the wake of generative AI, two things matter: being clever, and having a dope perspective - having good taste. Point of view. Yours. Your team's. Your agency's. Irreplaceable. But only if you can encode it into something that doesn't evaporate when someone walks out the door. When everything your agency knows feeds everything it does, and every project makes the next one smarter, the energy stops dissipating. It compounds. The arc gets wider with less effort. We call it ResonanceXL. First cohort is forming. Limited spots. If you know an agency owner who needs to see this, share it. If you want the little guy to stay in the fight, drop a reaction. If you think none of us should touch Gen AI, write a lengthy comment about how we're destroying society, share it far and wide, and then scroll on because we aren't talking to you.
-
Nicholas King liked thisNicholas King liked thisAfter 13+ years with Flooring Studio, I’ve stepped out and am now setting up The Flooring Room in Auckland. We’re getting the space ready for a 20th of April opening, a little painting, a bit of construction, and pulling everything together. Still plenty to get through before we open.
-
Nicholas King liked thisNicholas King liked thisThere's a specific kind of tired that feels really good. Deep Vibe just came out of one of the most varied stretches of work I can remember: 3D scenes and dramatic reveals for a confidential large-scale event (what happened in Vegas stayed in Vegas this time), four brand films, logo animations, a 3D animated brand anthem, and a documentary project that grounded everything. And another confidential one... a project that pushed our entire team deep into creative AI workflows, solving for scene and character consistency at a level not many have tackled before. The kind of work that expands what you thought was possible. We've been traveling a lot. We're heading back to Las Vegas next week and then straight to Denver. And honestly? I can't wait to see what the rest of this year holds. We'll share more as we can. Stay tuned! 👀
Experience & Education
-
Data Kinetic
View Nicholas’s full experience
See their title, tenure and more.
or
Recommendations received
3 people have recommended Nicholas
Join now to viewOther similar profiles
-
Melissa Womack
Melissa Womack
Birlasoft, recognized as India's Most Admired & Valuable Power Brand Company 2020
Explore more posts
Explore top content on LinkedIn
Find curated posts and insights for relevant topics all in one place.
View top contentOthers named Nicholas King
1410 others named Nicholas King are on LinkedIn
See others named Nicholas King