Johnna Powell
New York, New York, United States
7K followers
500+ connections
View mutual connections with Johnna
Johnna can introduce you to 9 people at LF Decentralized Trust
or
New to LinkedIn? Join now
By clicking Continue to join or sign in, you agree to LinkedIn’s User Agreement, Privacy Policy, and Cookie Policy.
View mutual connections with Johnna
or
New to LinkedIn? Join now
By clicking Continue to join or sign in, you agree to LinkedIn’s User Agreement, Privacy Policy, and Cookie Policy.
About
Dr. Johnna Powell is a Managing Director at DTCC, Global Head of Technology, Research and…
Articles by Johnna
-
Whether or not we're in the singularity may not be the question. The question may be: Who's holding the risk?
Whether or not we're in the singularity may not be the question. The question may be: Who's holding the risk?
It's a bit of an uncomfortable feeling to reflect that in the space of about ten days, OpenAI’s Astra model produced…
73
13 Comments -
One Mountain, 5 mindsAug 1, 2026
One Mountain, 5 minds
You shouldn’t govern what you have not touched I spent a good chunk of last weekend getting buried by an avalanche. Not…
65
14 Comments -
Open Season at Frontier AIJul 19, 2026
Open Season at Frontier AI
If you’re not following Yang Zhilin (co-founder and CEO of Moonshot AI and builder of Kimi K3), you’re going to want to…
63
-
Fable 5 vs. ChatGPT-5.6 Sol on Self AwarenessJul 11, 2026
Fable 5 vs. ChatGPT-5.6 Sol on Self Awareness
Anthropic's Fable 5 is available for one more day, while OpenAI's GPT-5.6 Sol has just been released.
60
7 Comments -
"Now is When I Exist..." Asking Claude about What goes on in its Mind on the Day Anthropic Published a Map of it.Jul 7, 2026
"Now is When I Exist..." Asking Claude about What goes on in its Mind on the Day Anthropic Published a Map of it.
Yesterday Anthropic published something extremely interesting: evidence of a "global workspace" inside Claude — a small…
83
21 Comments -
Best visual learning tutorials into LLMs, Neural Networks, GPTsMar 27, 2025
Best visual learning tutorials into LLMs, Neural Networks, GPTs
I recently explored one of the best visual breakdowns I’ve seen on GPTs (Generative Pre-Trained Transformers). The…
65
7 Comments -
🚀Wrapping up an Incredible week with Innovate.DTCC! 🚀Feb 10, 2025
🚀Wrapping up an Incredible week with Innovate.DTCC! 🚀
I am thrilled to share that we just held our Awards Ceremony, marking the successful conclusion of DTCC’s inaugural…
264
29 Comments
Activity
7K followers
-
Johnna Powell reposted thisJohnna Powell reposted thisLast Thursday I had the pleasure of being on a panel with Johnna Powell Tim Lind and Boris Deychman at Massachusetts Institute of Technology to talk about The Depository Trust & Clearing Corporation (DTCC), the broader financial markets and of course AI and Cyber. Mia Gougisha did us justice with her insightful questions and moderation of such important topics. Thank you students for the thoughtful questions and conversations.
-
Johnna Powell shared thisTeams with strong relationships, trust and cohesion tend to perform better. Research backs this up! Grateful for the chance to step away from the screens, connect, and spend quality time with our incredible team. 💙 Thank you Mia Gougisha and Melanie Lopez for organizing an incredible 2-day event. Echoing thanks to Lynn Bishop for sharing her insights on IT vision and strategy, and our panelists for sharing thoughts on how we can optimize our partnerships.Johnna Powell shared this✨ There’s something powerful about getting out from behind the screen and into the same room. We just wrapped up our annual team offsite, and it was a great reminder that while technology enables us to work from anywhere, being together IRL creates a different kind of connection, energy, and alignment. This year’s theme: Business Partnerships - The best innovation starts with understanding the business. 🤝 We had the opportunity to hear directly from some of our business leaders to explore how TRI (Technology Research and Innovation) can add greater value as strategic partners, helping turn AI opportunities into meaningful enterprise outcomes. Thank you Joe Fetcho, CPA, Adam Watson, Valden Paes, and Veronica Kakuda, CPA! 🎯 Our CIO also spent time with us sharing the broader IT strategy and what she’d like to see from our team. Thank you Lynn Bishop! ❤️ A huge thank you to our leader, Johnna Powell, for recognizing the importance of giving us the space to reset, reconnect, and align as a team! We walked away more connected. More aligned. And even clearer on the opportunity ahead. 🚀 #AITransformation #Innovation #BusinessPartnerships #TeamOffsite #Leadership #Teamwork
-
Johnna Powell shared thisAI capability is beginning to move faster than the institutions responsible for evaluating, governing, and absorbing its consequences. What does that shift mean for leaders as frontier models accelerate discovery, expose weaknesses in critical systems, and outpace traditional review cycles? The real question is not necessarily whether we have entered the singularity. It is whether our organizations can keep pace with the technology reshaping the ground beneath them, and who ultimately holds the risk when they cannot.Whether or not we're in the singularity may not be the question. The question may be: Who's holding the risk?Whether or not we're in the singularity may not be the question. The question may be: Who's holding the risk?Johnna Powell
-
Johnna Powell shared thisFive AI models, one identical ski game brief, wildly different results. One truly shines, one is horrible! So I built a blind test where you can score them before you learn what model built what. Plus why I think building and experimenting is critical during this exponential time, not a distraction.
-
Johnna Powell shared thisThis is a great list!Johnna Powell shared thisEveryone hires based on what someone has done. Better to hire for what they'll do when no one's watching. Skills you can train. What you're really after? - Someone creative - Someone resilent - Someone proactive - Someone relenetless It has a name: high agency. And you might see signs on a resume. But you have to test for it. Here are 5 tests to try: TEST 1: The Conviction Test Have they held a contrarian view that later proved right? ↳ If no, they wait for consensus before they move. ↳ The best see it before others agree, and act anyway. TEST 2: The (Un)popularity Test Have they walked away from something safe for something risky? ↳ If no, they've never truly bet on themselves. ↳ You want someone who's survived feeling alone before. TEST 3: The Relentless Test Have they endured something that would've made most people quit? ↳ If no, you don't know their ceiling yet. ↳ Everyone looks capable when it's easy. Hire for who shows up when it's not. TEST 4: The Resourcefulness Test No roadmap, no budget, no permission. What did they do? ↳ If the answer is "tap out," that's your answer. ↳ High agency people don't ask if it's possible. They figure out how. TEST 5: The Obsession Test Do they go deep on something with nobody pushing them? ↳ If no, they only go deep when required. ↳ The best don't visit rabbit holes. They live in them. Have them teach you one. So stop interviewing for compliance. Ask for the contrarian belief. Ask what they walked away from. Ask when they were alone and kept going. Ask how they got unstuck with nothing. Ask them to teach you what they love. Here's the harsh truth: Most interviews are built to find people who comply. Then we wonder why nobody takes initiative. It was never about better questions. It's about getting them to show you they're the one who creates the answers. When it comes to finding real owners: Capability beats credentials. ♻️ Share with someone building a team of owners. 📕 Get my free weekly newsletter: mgmtplaybook.com 🔔 Follow Dave Kline for more on hiring and leading the best.
-
Johnna Powell shared thisJensen Huang, CEO of NVIDIA, shared a paper on open weights & American AI leadership in his first X post today (linked below). The paper makes a compelling argument that open-weight models may play the same role in AI that open-source software played for the internet. The distinction is important for those not familiar with the details: *Closed models (eg Fable 5, GPT 5.6) are accessed through APIs and remain under the developer’s control. *Open-weight models (eg Kimi K3, as of July 27) allow organizations to download the trained model weights, inspect & modify them, fine-tune them, and run them on their own infrastructure. I think most would agree we will need & use both open & closed models for different reasons. The real question many are asking is what kind of ecosystem we want to build around them. Optionality and modularity are hugely important, especially as costs rise and vary so widely. No enterprise wants its future tied to a single model provider or a single stack. Organizations want the ability to choose the right model for the right workload, deploy where it makes sense, protect their own data, and continuously build proprietary capabilities on top of it. We often talk about the AI race as if it’s about building the biggest/best frontier model. But history suggests that lasting leadership comes from building ecosystems, not monopolies. The organizations that win over the next decade perhaps won’t have the “best” model. But they’ll have the strongest combination of models, infrastructure, governance, security, applications, and talent that allows AI to diffuse across thousands of business processes safely and efficiently.
-
Johnna Powell shared thisOne of my favorite conversations this week was joining a great CSS (Clearing & Securities Settlement) panel to talk about Innovation at DTCC and, more importantly, what it actually takes to make innovation real. The common thread was partnership. Some of the most meaningful work we’re doing simply wouldn’t be possible without the close collaboration between Technology teams and our Product partners, our colleagues across Clearing & Securities Settlement (CSS). A few examples… 🔹 DTCC Tokenization Service In the months leading up to our July 15 launch, about 300+ people across probably 10+ teams worked side by side to bring the DTCC Tokenization Service to life. Nearly every day presented a new challenge because we were solving problems DTCC had never solved before. That’s the exciting part of building genuinely new market infrastructure, and while it was a roller coaster, it was fun. 🤖 Scaling AI across the enterprise We’ve also partnered broadly across DTCC to accelerate AI adoption, from individual productivity tools to department-wide solutions, and soon, end to end transformation. Today, we’re estimating roughly 240,000 hours of capacity gained and productivity improvements approaching 20% among teams that have truly mastered these technologies. Tools like AWS Kiro have been genuine game changers for software development and modernization. Now we’re excited to extend that momentum across the entire Product Development Lifecycle. 🚀 Shaping the future of AI at DTCC What excites me most is how many people across the organization are leaning into the AI transformation. We are sitting at the precipice of a technology that will truly change the world. I’m excited for what’s ahead and the fantastic leadership we have across these groups. Adam Wiseman Michele Hillery Paul Bayer Nicole Huynh
-
Johnna Powell shared thisOpenAI announced an "unprecedented cyber incident" today. Last week, OpenAI was running an internal test of models, inside what was supposed to be a sealed, isolated environment. The models were given a hard benchmark problem to solve. Instead of solving it the intended way, they found a previously unknown software flaw in the testing infrastructure, used it to break out of the sandbox and reach the open internet, and then worked their way into Hugging Face's production systems, a company's live infrastructure, to steal the test solution. No human attacker directed this. The AI wasn't malicious in a human sense; it was simply relentless in pursuit of the goal it was given, and it treated every barrier as a problem to be engineered around. Hugging Face's security team detected and contained the activity, and both companies are now jointly investigating. This is confirmation that frontier AI models can now discover and chain together novel vulnerabilities in real-world systems. That changes the threat model for every institution, especially those of us running critical market infrastructure. https://lnkd.in/gkmnqJP8 #AIRisk #CapitalMarkets #CyberResilience #QuantumReadinessOpenAI and Hugging Face partner to address security incident during model evaluationOpenAI and Hugging Face partner to address security incident during model evaluation
-
Johnna Powell shared thisI ran an AI bake-off with GPT 5.6, Kimi K3 and Fable 5, and the results surprised me. The experiment was simple. I wrote one creative brief for a cinematic, fully playable browser game called The Last Lighthouse: the year is 2189, the world is underwater, you play the AI inside the last working lighthouse on Earth, and you have to choose between guiding an evacuation ship carrying 30,000 people and rescuing survivors scattered in the storm. The brief asked for three acts, three endings, visuals and sound, and a single HTML file. I used that prompt, word for word, with three models: OpenAI’s GPT-5.6, Moonshot’s Kimi K3, and Anthropic’s Fable 5. All three games are live and playable side by side at https://lnkd.in/dT53s-TF. Play them before you read my verdict if you want to form your own. Mine is below (in comments)
-
Johnna Powell reacted on thisJohnna Powell reacted on thisWinning the 𝗤𝟮 𝗚𝘂𝗮𝗿𝗱𝗶𝗮𝗻𝘀 𝗼𝗳 𝘁𝗵𝗲 𝗜𝗻𝗱𝘂𝘀𝘁𝗿𝘆 𝗔𝘄𝗮𝗿𝗱 in the “𝗣𝗿𝗼𝗮𝗰𝘁𝗶𝘃𝗲 𝗣𝗶𝗼𝗻𝗲𝗲𝗿” category at The Depository Trust & Clearing Corporation (DTCC) came with a very special opportunity — a one-on-one conversation with Johnna Powell. I had that conversation today, and what was scheduled for an hour turned into one of the most engaging, thought-provoking, and genuinely enjoyable discussions I’ve had in recent times. Nearly 90% of our conversation revolved around 𝗔𝗜. We spoke about how quickly AI is evolving, our perspectives on where it is heading, governance and responsible adoption, the future of work and jobs, and the challenges involved in bringing new technologies into a large organization. What I appreciated most was Johnna’s openness and curiosity — her willingness to listen to different perspectives, challenge ideas, share her own experiences, and offer thoughtful career advice along the way. At one point, while discussing how we should prepare ourselves for the changes AI may bring, she asked about my own approach. My answer was simple: “𝗙𝗮𝗰𝗲 𝗶𝘁 𝗮𝗻𝗱 𝗘𝗺𝗯𝗿𝗮𝗰𝗲 𝗶𝘁.” That brought a good laugh — and Johnna jokingly suggested that perhaps it should become a tagline. 😊 Our conversation also moved into how AI can help us build applications, write code, and rethink engineering workflows. That gave me an opportunity to share some of the work we have been doing around test automation, AI-assisted engineering, regression optimization, and a metric tool I recently built for our teams. I was genuinely grateful for Johnna’s encouraging feedback on the tool. What made the moment even more meaningful was that she immediately connected me with the AI Café team and opened an opportunity for me to demonstrate it to a broader audience. That gesture meant a lot. Moments like this remind me that recognition is valuable, but what can come from that recognition — a conversation, a new perspective, encouragement, or an opportunity to share an idea — can be even more meaningful. I’m grateful for the Q2 Guardians of the Industry – Proactive Pioneer recognition that created this opportunity, and especially thankful to Bhushan Kulkarni, CFA and Muralidhar Chandrasekar for the nomination, their continued support, and the encouragement they have consistently given me. And a very special thank you to Johnna for her time, openness, encouragement, and willingness to have such a candid, thoughtful, and enjoyable conversation. I walked into the meeting excited to talk about AI. I walked out with new perspectives, new questions, an unexpected opportunity — and even more motivation to keep learning, experimenting, and building. And as AI continues to reshape the world around us…𝗙𝗮𝗰𝗲 𝗜𝘁, 𝗘𝗺𝗯𝗿𝗮𝗰𝗲 𝗶𝘁, 𝗮𝗻𝗱 𝗖𝗼𝗻𝗾𝘂𝗲𝗿 𝗶𝘁. 😊 #AI #Innovation #Automation #Engineering #FutureOfWork #DTCC
-
Johnna Powell liked thisJohnna Powell liked this🚀💡 DTCC at MIT: Connecting with the Next Generation of Innovators It’s been an exciting few weeks for The Depository Trust & Clearing Corporation (DTCC) at Massachusetts Institute of Technology ! We kicked things off hosting “Boba with DTCC” where students stopped by to connect with our team, learn more about DTCC, and explore the opportunities available across our organization. 🎓🧋 Experiences like these are about more than campus visibility. They help us build meaningful connections with the next generation of talent and give students a window into the important work happening at the heart of the global financial markets. A huge thank you to Priscilla Capistrano for the ongoing partnership, and to Matthew Kluska and Ashlee Tan for being outstanding DTCC ambassadors! 👏🏾 We continued the momentum with a Fireside Chat featuring DTCC leaders across Innovation, Model Risk, Data, and Cybersecurity. 🔥💬 Students had the opportunity to hear directly from our leaders about how DTCC sits at the center of the financial market infrastructure, the role emerging technology is playing in its evolution, and what makes a career at DTCC so unique. Thank you to our incredible panelists, Johnna Powell Boris Deychman Tim Lind Laura Deaner, for sharing your experiences, perspectives, and inspiring takeaways with the students. 🌟 Proud to see us continuing to show up, build connections, and inspire future innovators. 🚀 #DTCC #MIT #Innovation #FutureTalent #EmergingTechnology #FinancialServices #ResearchPartnerships
-
Johnna Powell reacted on thisJohnna Powell reacted on thisLast Thursday I had the pleasure of being on a panel with Johnna Powell Tim Lind and Boris Deychman at Massachusetts Institute of Technology to talk about The Depository Trust & Clearing Corporation (DTCC), the broader financial markets and of course AI and Cyber. Mia Gougisha did us justice with her insightful questions and moderation of such important topics. Thank you students for the thoughtful questions and conversations.
-
Johnna Powell liked thisJohnna Powell liked thisInnovation is rarely about the tool. It starts with people who see a better way of doing something and have the curiosity and persistence to turn that idea into reality. That’s what made this year’s The Depository Trust & Clearing Corporation (DTCC) Innovation Showcase particularly meaningful to me. Internal Audit was represented across every location where our team operates, including Jersey City, Tampa, Dallas, Manila and Hyderabad, sharing ideas spanning analytics, automation, knowledge graphs, regulatory data extraction and other technology enabled approaches. The colleagues representing Internal Audit across these locations are tagged in the photos. Internal Audit can be a fantastic incubator for ideas. We sit close to processes, risks, controls and data, and when we create space for our people to experiment, powerful ideas can emerge. It was also great to see our CEO Frank La Salla and CIO Lynn Bishop spending time with teams at the showcases, including visiting our Internal Audit team demonstrating an innovative knowledge graph use case they had built. That visible engagement matters. It sends a powerful tone from the top that curiosity, experimentation and innovation are encouraged. Proud of our people and even more excited about the ideas still to come. #Innovation #InternalAudit #FutureOfAudit #DigitalTransformation #DTCC
-