Cassie Kozyrkov

Kozyr

Speaking and advisory services. Request form: makecassietalk.com

Kozyr was renamed from Data Scientific.

Chief Executive Officer

Federal Reserve Bank of New York

Served a full two year term as Innovation Advisor and Councillor at the Federal Reserve Bank of New York, with a focus on the impact of AI on the financial markets and global policy.

Innovation Advisory Council Member

Google

Decision Intelligence is an interdisciplinary field concerned with all aspects of decision-making.  It combines Data Science (statistics, machine learning, AI, analytics) with the Behavioral Sciences (psychology, neuroscience, economics, and the managerial sciences).

I serve as a thought leader in this new discipline, guiding Google and the world in making safe, effective use of AI and data.

Chief Decision Scientist

My mission is to democratize artificial intelligence and guide business leaders in harnessing the transformative power of machine learning.

Chief Data Scientist, Office of the CTO

Youngest ever member of Google's Office of the CTO, specializing in AI and data science.

I'm also a coauthor of Google's Technical Director training program to teach others how to succeed in the role.

Technical Director, AI and Data

My roles are: statistician, machine learning expert, mentor, data scientist, evangelist, decision scientist, and consultant.  I bridge the gap between leaders and quantitative analysts / researchers.

As Google's statistics and machine learning sensei:
* I've advised 200+ teams at Google on their analytics processes.
* My ideas have sparked 400+ new projects.
* I've trained 20,000+ analysts/leaders/engineers in rigorous data-driven decision-making and/or using machine learning to transform their work.

Largest projects in: Search Advertising, Google Maps, Machine Intelligence, Infrastructure, YouTube

Statistician

Self-Employed

Data science consulting and analytics services for clients from government organizations, private sector firms, and academia.

Data Science Consultant

North Carolina State University

Statistics Lecturer

Statistical Computing, Advanced Topics in Statistics: Big Data, Special Topics in Statistics: Data Mining, Applied Time Series Analysis, Applied Multivariate Data Analysis, Advanced Probability and Measure Theory, SAMSI Computational Advertising, Statistical Methods II, Statistical Consulting, Statistical Theory I, Statistical Theory II, Linear Models and Variance Components, Stochastic Processes, Doctoral Research, Supervised Doctoral Teaching, R Programming

Mathematical Statistics PhD Student

Duke University

Psychology & neuroscience with specialization in transcranial magnetic stimulation (TMS) methods and the science of decision-making.

Neuroeconomics PhD Candidate

Center for Cognitive and Social Neuroscience at the University of Chicago

Coordinated lab research projects and team effort. Designed and ran experiments.

Project Manager

Charitable contributions and decision-making research with Professors Steve Levitt and John List.

Research Assistant - Prof. John List

Cassie is Google’s former Chief Decision Scientist and a sought-after AI advisor, working with senior leaders at places like Lenovo, NASA, and Gucci. She is currently CEO of Data Scientific and is known for founding the field of Decision Intelligence.

Cassie is a beloved personality in the data and AI community, followed by over half a million tech professionals. She is also a renowned keynote speaker and writer. ( If you'd like her to speak at your event, here's her booking form: makecassietalk.com )

In almost 10 years at Google, Cassie personally trained over 20,000 Googlers in data-driven decision-making and AI and has helped over 500 projects implement decision intelligence best practices. Before that, she served in Google's Office of the CTO as Chief Data Scientist. The rest of her 20 years of experience was split between data science, consulting, lecturing, and academia.

If you've had a good chuckle while learning about AI, statistics, or decision-making, chances are you've encountered her writing, which has reached millions of readers. She has been honored as a LinkedIn Top Voice for 5 years and as a #1 Artificial Intelligence writer on Medium for 7 years.

Speaking and advisory services. Request form: makecassietalk.com

Gemini 3 is here and for the first time in years, Google is not chasing the frontier... it *is* the frontier. 

In the last 48h, Google has leapfrogged all the other frontier models.

And that's not all:
💪  Nearly half of Google’s new code is now AI-generated and engineer-reviewed. A sign that Google is already operating at a different internal velocity. 
💪  Google's Ironwood TPU delivers 42.5 exaflops — about 24× the power of the world’s largest supercomputer (El Capitan).
💪  Gemini Enterprise shows a full-stack strategy: model + infra + agents + security + workflow integration, all under one roof.
💪 Antigravity, Google's Agentic IDE, is here. It's designed for an “agent-first” workflow where AI agents actively operate across the editor, terminal, and browser.

Google hasn’t just rejoined the race — it’s out in front, and the rest of the ecosystem now has to respond. And they are responding. 

Nvidia, Microsoft, OpenAI, and the world-model labs each dropped their own shockers — and all but one are industry-defining in their own right. Can you guess who the laggard is before you read the deep dive in my newsletter? 

Newsletter: decision.substack.com 

Before you look it up, put your guess below. 👇 After all, your comments are how posts get seen and if you liked this one, it's a lovely way to help others enjoy it. 

Your comments make the (digital) world go around and your reposts ♻️ make my day.

For a full picture of how every AI behemoth just repositioned itself, my analysis is here: bit.ly/quaesita_behemoths

Don't forget to tap that follow button for more (especially if you don't want to miss my post on AI in China next week).

#ai #google #nvidia #openai #microsoft

P.S. My Decision-Making with ChatGPT course runs on demand next week with a 2h live Q&A. Enroll here: https://lnkd.in/eG7EBtTv (for promo codes, head over to the newsletter article 👆)

CEO, Google's first Chief Decision Scientist, AI Adviser, Decision Strategist, Keynote Speaker (makecassietalk.com), LinkedIn Top Voice

🤦  🤦  🤦  While 83% of companies plan to deploy AI agents, only 13% have the foundations to make them work. 🤦  🤦  🤦 

The rest? They’re about to learn an expensive lesson about the difference between ambition and capability.

Despite all the investment, all the pilots, all the executive enthusiasm, we’re watching the market split into two distinct camps: the prepared and everyone else. 

My newsletter digs into what separates them, link in the first comment. 

#aiagents #ai #agents

For added treats, it announces my new course(Decision-Making with ChatGPT) and included promo codes subscribers can use for 25% off.

As always, your comments and reposts make my world go around.

For those who haven't had a chance to enjoy my 85min course on decision intelligence yet... 

I've toggled the friendly button for you for a couple of days so you can help yourself to it free* with the link below. 

Enjoy!

#datadrivendecisionmaking
#decisionmaking

*Well, not free-free, because my price is that you pretty please show the course some love if you get value out of it it and use the opportunity to share it for free with your community too.

AI risk isn’t someone else’s job.

Too often, leaders assume it’s the domain of IT, compliance, or data science... but the truth is, AI risk cuts across every part of the business. Brand, trust, liability, reputation are all on the line.

Here are five categories of AI risk every executive needs to own:

1️⃣ Data security: AI expands your attack surface in ways traditional systems never did.
2️⃣ Output quality: “Polished” ≠ “reliable.” Verification has to be part of the workflow.
3️⃣ Bias: Every dataset reflects human choices. Leadership decides whether to amplify or correct.
 4️⃣ Privacy: AI can expose patterns you didn’t even know were sensitive.
 5️⃣ Compliance: Global scrutiny is rising fast; reactive mode isn’t a strategy.

Each of these connects directly to organizational trust and accountability. They’re not technical footnotes — they’re leadership imperatives.

Your turn: what other category of AI risk do you think belongs on this list? 

Your comments make the (digital) world go around and your reposts ♻️  make my day.

#AILeadership #DecisionIntelligence #GenerativeAI #AIinAction
#Strategy #DigitalTransformation

Don't forget to mash, ahem no, daintily tap that follow button for more.

Leaders are discovering the hard way that “plug-and-play AI” is mostly a myth. The tech may look polished, but the real work begins where the glossy demo ends.

This new episode of AI in Action breaks down why so many teams burn budget, time, and momentum before they ever see meaningful value — and why the missing ingredient is almost always the same: precise, upfront definitions of success.

❌ Clear metrics aren’t paperwork. 
✅ They’re strategic infrastructure. 

Without them, AI initiatives drift, stall, or quietly accumulate hidden costs. 
With them, teams avoid rework, manage risk, and actually ship systems that hold up under the pressure of real users and real workflows.

If you’ve ever watched an AI project spiral because “it looked easy at first,” this one will resonate. We also dig in to an irreconcilable conflict between you and your vendor.

🎥 Video below.

Your turn: What’s one “plug-and-play” promise you’ve seen turn into a long, messy integration story?

Your comments make the (digital) world go around and your reposts ♻️ make my day.

#AILeadership #DecisionIntelligence #GenerativeAI #AIinAction
#Strategy #DigitalTransformation

Don't forget to mash, ahem no, daintily tap that follow button for more.

Whoops, did I miss the toddler's birthday yesterday?? 

Happy belated 3rd birthday, ChatGPT, you little weirdo! 🎂 🥳

In the AI-first era, intelligence is the baseline -- leadership is the edge.

When intelligence is everywhere, technology alone isn’t what sets organizations apart. The real advantage comes from how leaders connect data, insight, and action to make decisions that move the business forward.

That’s what I explore in my keynote on AI-first leadership, looking at what it takes to guide organizations in the AI era, where the advantage comes from connecting human judgment with data-driven insight, and where clarity and accountability define success.

Because the future of AI isn’t just about smarter machines. It’s about smarter decisions, and the leaders ready to make them.

👉  Watch the keynote → https://lnkd.in/eMJhFQ5Y
 
#UnleashAI #ReimagineCX #createaNiCEworld #sponsored

If you like to use the Thanksgiving break to get smarter about what’s actually happening in #AI, this month delivered a feast. Come for the #geopolitics, stay for the robot comedy.

A few highlights from this month’s global state of play covered in my newsletter:

🔴 Washington reached for the Manhattan Project analogy.

In its own memo, the White House described its new AI initiative as “comparable in urgency and ambition to the Manhattan Project.”
And in case you missed it, the US is restricting advanced chip imports to China... or are they? The drama continues!

🔴  China accelerated its push for AI self-sufficiency.

Beijing banned foreign AI chips from state-funded data centers and rolled out power subsidies of up to 50% for firms adopting domestic hardware. Baidu unveiled plans for two new processors plus “supernode” clusters linking hundreds of its Kunlun chips to mimic Nvidia-class performance while supporting Nvidia’s CUDA ecosystem. China isn’t waiting for imports; it’s rebuilding the stack.

🔴 Europe is considering a GDPR ease-up.
 
Yes, really.
 Broader “legitimate interest,” simpler reporting, fewer things counted as personal data. A rare philosophical shift from Brussels.

🔴 The Middle East keeps accelerating. 

Their enterprises spend close to 10% of revenues on digital transformation with faster ROI cycles than the US or EU. The UAE is even exporting AI infrastructure into Africa.

🔴 The US regulatory picture is lose–lose.
 
A tightening patchwork of state laws collides with a federal push for preemption. Prepare now and risk wasting the work; prepare later and risk penalties. No good options.

🔴 Radical transparency arrived out of nowhere.
 
AI2 released a 32B-parameter reasoning model with full weights, checkpoints, datasets, and training code. Unusual clarity in an increasingly opaque field.

🔴 Robots!

One country gets back-to-back (or should I say face-to-floor) robotics embarrassments, while another sets a Guinness World Record and has to cut of its humanoid bot's leg to prove it isn't a human in a suit. Can you guess the two countries?

🔴 Leaked memos and more...

Including:

An internal OpenAI memo admits “economic headwinds” now that Google has the crown again.
A reveal of how many data centers Amazon actually has (analysts have been way off).
Claude's release 2 days ago, Opus 4.5, outperformed every performance engineer Anthropic has ever tested, human or otherwise.

What a month.

If you want the full breakdown—the context behind each shift and what it means for the months ahead—it’s all in this today's newsletter.

Here’s the link again: decision.substack.com

💌 If you found this useful, a repost or comment helps more people walk into Thanksgiving conversations actually knowing what’s going on in AI. 

And honestly, if you enjoyed it, helping my writing reach someone who would find it useful is the best thank you. 🙏

My new course, Decision-Making with ChatGPT --  https://lnkd.in/eG7EBtTv -- got such glowing reviews, but there's one huge problem with it... 😬

It's for leaders and aspiring leaders and curious self-improvers. In other words, the busiest people on earth! I asked Cohort 1 to carve 3 hours out of their impossible schedule for an all-live virtual course with me.

I hear you! And I know some of you want to watch everything on 3x speed -- you prefer me with a proper chipmunk voice, I get it. 🐿️🐿️🐿️

I have taken your feedback and created a new HYBRID version of the course! 

👉 Here's how it works:

SELF-PACED PART: The course portal opens today at 12 PM ET. You have all week to self-study 5 modules with 30+ edited videos at your own pace.

LIVE PART: Then come hang out with me live for 2 hours and bring all your burning questions so we can go deep and cement your understanding.

Depending on your taste for homework and study guides and exercises, you can complete the on-demand portion in 2 hours (sped up minimalist version) or 8+ hours (maximalist version). Choose your own adventure for something in between and I'll see you soon! Or go fast now and slower later, you'll have lifetime access.

Enroll here: https://lnkd.in/eG7EBtTv

Hope to see you there!! And if you can't join but you like the initiative, repost and comment for reach for the course finds its way to those who need it.

PROMO codes: You have two choices for how you'd like to earn a discount.

1) Subscribe to my newsletter decision.substack.com (free or paid, up to you) to take $200 off. Code: SUBSCRIBERS.

OR

2) Commit to being a champion of the course to take $300 off. Before the course, that means helping the course find its way to at least 5 people who you think would benefit from it. After the course, it means putting the effort in to leaving an honest review on social media. No inflation though! Honesty matter. Code: CHAMPIONS

One code per person, your choice.

#ai #decisionmaking #chatgpt

🦾  Calling all long weekend self-improvers! 🦾 

Last chance to take my Decision-Making with ChatGPT course at a time that's definitely not going to clash with work hours...

Enroll here: https://lnkd.in/eG7EBtTv

The course is running on-demand right now and the live component, a 2 hour Q&A opportunity with me, is tomorrow afternoon from 1 PM to 3 PM ET.

Will I see you there? Drop me a comment or repost to make my day. 😊 

Deal hunters can find promo codes in here: https://lnkd.in/eXkjb-da

Over 1.2 MILLION of ChatGPT’s weekly active users engage in conversations that contain explicit indicators of suicidal ideation. (Source: OpenAI)

OpenAI tells us:
GPT-5 shows 🟢safe🟢  behaviors in 91% of suicidal conversations.

Which also means that:
GPT-5 shows  🔴unsafe🔴 behaviors in 9% of suicidal conversations.

These are both big numbers, all things considered. 

Is ChatGPT a dangerous substitute for genuine human connection and professional help with a 9% failure rate (roughly 108,000 people per week), or is it a valuable and uniquely accessible first line of support for those in distress?

In this week's edition of my newsletter, let’s talk about robo-therapy. Everything from the pros and cons to the FDA's decision about whether to allow GenAI for mental health.

👉  https://lnkd.in/eG9BRDbt

What are your thoughts? Would you want AI therapy to be legalized?

#AItherapy #ai #mentalhealth

If you found this useful, a repost makes my heart happy. 💌  And a subscription to my newsletter makes my day: decision.substack.com

(Or share this with the smartest person you know.)

How many of these AI-powered scams are you familiar with?

🐡 Chatfishing
🐷 Pig butchering
📣 Voice cloning

Any stories to share that you or a friend experienced?

Not sure what any of these are about? Let me walk you through them in my newsletter: https://lnkd.in/eEvxKs7h

#aiscams #ai #security

If you found this useful, a ♻️  repost makes my heart happy. And a subscription to my newsletter makes my day. 💌

Open article: 3 New AI Scams You Need To Prepare For by bit.ly

How many of these 9 #AI #headlines did you catch recently? Tell me the one you find most interesting in the comments 👇 

For the curious, my newsletter digs into each one: decision.substack.com 🔗  (link in comments)

🗞️ 1 Nvidia becomes the world’s first $5 trillion company
🗞️ 2 AI anxiety drains employee motivation
🗞️ 3 AI is 50% more sycophantic than people
🗞️ 4 $20K home robot isn’t as autonomous as it looks
🗞️ 5 OpenAI unveils Aardvark, an AI-powered security researcher
🗞️ 6 Scale AI unveils benchmark to measure how much work AI can really do
🗞️ 7 Google AI debuts “Supervised Reinforcement Learning” for small models
🗞️ 8 Wharton: Gen AI Has Officially Gone Mainstream in Business
🎤  9 AI artist Xania Monet debuts at #30 a Billboard radio chart

More info in my newsletter! (link in comments)

Don't forget to tell me if one of these headlines is more interesting to you than the others so I know what to look for to bring you more tasty things next week. 

If you found these #ai #news updates #useful, words of affirmation are my love language. :) As are reposts! ♻️

😂 For those who missed some peak unintentional comedy last week...

Built to demonstrate AI independence amid sanctions with 77% domestic components, Russia's first AI-powered humanoid robot faceplanted at its grand debut to the soundtrack of Rocky's Gonna Fly Now.  Lyrics:

Trying hard now
It's so hard now
Trying hard now

Oh, he's ready
Gonna make a move, yeah
Oh, he's ready
And makin' moves, yeah

Getting strong now
Won't be long now
Getting strong now

Gonna fly now
Flying high now
Gonna fly, fly

Happy Sunday!

https://lnkd.in/e5yEYPUk

Many organizations are discovering that the real leverage point in the AI era isn’t the model — it’s the leader. 

As systems take on more decisions at scale, the consequences of unclear intent, weak guardrails, or fuzzy problem-framing multiply fast.

Companies keep investing in more powerful AI, yet the real differentiator is becoming unmistakable: the leaders who can guide decisions at scale.

When an AI system rolls out, it doesn’t just generate outputs — it creates a new layer of automated decisions that ripple through teams, workflows, and culture.

If leaders don’t define the context, the intent, and the guardrails, the system fills in the gaps on its own. That’s where misalignment (and costly surprises!) begin.

This new episode of AI in Action digs into why human judgment has become the defining variable in whether AI accelerates clarity or amplifies confusion. They’re the ones who understand how technical decisions, human decisions, and organizational decisions interact.. And who shape that ecosystem with intention.

The focus is on decision leadership: understanding how technical outputs, human choices, and organizational dynamics interact... and how to guide that ecosystem with precision. 

When leaders build this capability, AI stops being a risk magnet and becomes a force multiplier.

The result is a competitive advantage that scales: better framing, better choices, and better outcomes, amplified by AI rather than distorted by it.

👉 Check out the video and don't forget to share it with the smartest leader you know.

#AILeadership #DecisionIntelligence #GenerativeAI #AIinAction
#Strategy #DigitalTransformation

Your turn: Which layer is proving most challenging in your world right now — technical, human, or organizational? Your comments make the (digital) world go around and your reposts ♻️  make my day.

Happy Thanksgiving, everyone! Some 🦃  updates:

1) Want to win some arguments at tonight's family dinner? 

Here's my cheatsheet on AI geopolitics in my newsletter (US, EU, China, Russia, Middle East): https://lnkd.in/eXkjb-da

2) Want to win at shopping? 

If you combine the following 2 tools, you have a very solid way to deal hunt like a futurist:

🛒 The new Shopping Research feature in ChatGPT (good for finding products but bad at price accuracy)
🛒 The new shopping update to AI Mode in Google Search (good for checking prices) plus Google's new agentic feature which calls stores to verify inventory on your behalf (good for spending less time on the phone with stores, probably bad for the stores themselves)

3) Want to win at decision-making? 

Join my new course, Decision-Making with ChatGPT, which is currently on demand and is having its live 2h Q&A with me this Saturday afternoon at 1 PM Eastern time. Enroll here: https://lnkd.in/eG7EBtTv ...deal hunters will find discount codes neatly tucked into the AI geopolitics newsletter article linked in (1)

4) Want to win at making my day? Comment for reach and share this with someone who'll enjoy it. 💌 

📷 I took this photo in Uniclo Hong Kong last week. :D Now I'm back in the States with my family and wondrously jetlagged. Enjoy your Thank You Festival, everyone!

I enjoyed getting into both the technical and the inspirational sides of AI with JetBrains’s Kris Kang. I believe that the best use of AI isn’t about pinching yesterday’s pennies; it’s about doing the previously impossible. 

Sign up for the episode to hear me dispel some common misconceptions among leaders, share my thoughts on vibe coding, and give a little tough love to my fellow recovering statisticians.

https://jb.gg/b4e7ao

#ai #leadership #vibecoding #sponsored

What's the % of the world's AI computing power held by each of these players?

🇨🇳 China - ❓ %
🇺🇸 US - ❓%
🇪🇺 EU - ❓%
🌍 Everyone else - ❓%

Got guesses? 👇 

The answers to this and so much more in my Global AI State of Play:

Substack -- https://lnkd.in/eXkjb-da 
Medium -- https://lnkd.in/eU-vBqRi

#ai #geopolitics #aihardware

If you found this useful, comments and reshares make my world go round. And a subscription to my newsletter makes my day. :)

It's not too late! If you woke up today thinking "I'd like to upgrade myself today" and it's before 11 AM Eastern Time...

Then you still have time to absorb enough of the on-demand content from my Decision-Making with ChatGPT course to be ready for the 2 hour live Q&A with me at 1 PM ET today. (And if you like to speed up your videos to enjoy chipmunk voice, I'll let you do the math.) 

Enroll here: https://lnkd.in/eG7EBtTv

So if you're the type of person who likes to bulk-learn and you're feeling inspired today, it's not too late to join.

Alternatively! There's also a Thursday afternoon ET option (Dec 18) if you like to space out your learning.

PROMO CODES: can be found by scrolling to the bottom of bit.ly/quaesita_chipwars

P.S. I'll also be running private cohorts for corporate teams in mid-December. (Applications via makecassietalk.com) If you know someone who might like to arrange a training for their team, please share this with them and comment for reach.

taplio

Cassie Kozyrkov's Linkedin Analytics

Cassie Kozyrkov

Performance ScoreCassie Kozyrkov's LinkedIn Performance

What is Cassie talking about?

Who is engaging with Cassie

Cassie Kozyrkov's Best Posts (last 30 days)

Ready to grow your LinkedIn presence?

taplio

.css-6ytdur{background-image:linear-gradient(284deg, rgba(39, 43, 72, 1) 0%, rgba(33, 47, 183, 1) 100%);color:transparent;-webkit-background-clip:text;background-clip:text;}Cassie Kozyrkov's Linkedin Analytics

Cassie Kozyrkov

Performance Score.css-goub6s{font-size:var(--chakra-fontSizes-md);font-weight:500;color:var(--chakra-colors-gray-600);display:block;margin-top:var(--chakra-space-1);}Cassie Kozyrkov's LinkedIn Performance

What is Cassie talking about?

Who is engaging with Cassie

Cassie Kozyrkov's Best Posts .css-1emoo84{font-size:var(--chakra-fontSizes-md);font-weight:400;color:var(--chakra-colors-gray-600);}(last 30 days)

Ready to grow your LinkedIn presence?

Cassie Kozyrkov's Linkedin Analytics

Performance ScoreCassie Kozyrkov's LinkedIn Performance

Cassie Kozyrkov's Best Posts (last 30 days)