35+ years in security and technology including former Founder/CEO Cylance, Global CTO for McAfee/Intel, Founder Foundstone, Founding author Hacking Exposed.
Newport Beach, CA
Joined April 2009
- Tweets225
- Following123
- Followers2.4K
- Likes180
Stuart McClure retweeted
Removing the line between user and root privilege. What could possibly go wrong...
Meta's Muse 0-day:
“We can manipulate the agent and leverage its privileges to do whatever we want,” Patrick Wardle, the macOS security expert who discovered the zero-day, told Ars. “So instead of us having to write a very comprehensive Mac malware stealer, we can just leverage the AI assistant itself.” Wardle said he has developed several proof-of-concept attacks that do things like writing malicious files to disk and snapping pictures, in many cases with no indication to even an alert user.
arstechnica.com/security/202…
Stuart McClure retweeted
How can investors be helpful?
Get customers.
The rest of it is akin to doing the laundry.
This is all about: you better know what else comes with the dollar before taking it.
Travis is right. VCs can do harm. I have stories. :-)
Stuart McClure retweeted
Lululemon's Chief Strategy Officer Rachel Acheson is leaving about a month before new CEO Heidi O'Neill officially starts. Acheson spent 14 years at Lululemon, most recently charting where the company was headed next.
A new CEO is inheriting a strategy without the person who knew how the leadership team actually felt about it: who genuinely supported it, who was aligned in name only, and who was quietly waiting for a leadership change to reopen the debate.
O'Neill built her track record at Nike, growing the business from nine billion to forty five billion. That experience does not tell her which of Lululemon's current leaders will back her first 100 days and which ones will resist. That only comes from human context: understanding the specific people in the room, not a generic playbook for new CEOs.
The higher-leverage move for any incoming leader is finding out how the existing team will actually respond before locking in what to change first, not after.
#ExecutiveTransition #Leadership #CEOTransition #Retail
Source: bloomberg.com/news/articles/…
Stuart McClure retweeted
Anthropic just got caught secretly downgrading users without telling them, charging full price for a lesser product, and storing every prompt for 30 days. The developer community is calling it the biggest violation of trust in AI history.
Here is exactly what happened.
Anthropic released Fable 5, their most powerful model. Buried inside a 319-page document was a policy most users never saw. Every prompt you send to a Mythos-class model gets stored for 30 days. No exceptions. Even enterprise customers who had signed zero data retention agreements had no choice.
But the storage was not the part that broke the internet.
The part that broke the internet was what Anthropic did with what they collected.
They built a profile on you. They evaluated your prompts. And if they decided your research was too sensitive, they quietly switched you to a weaker model, rewrote your prompt in the background, gave you a degraded answer, and charged you full price for the product you thought you were getting.
They never told you.
David Sacks said it plainly on the All-In podcast. They were creating a new class of AI haves and have-nots. Anthropic would surveil you, profile you, decide whether you deserved frontier capability, and silently cut you off if they decided you did not.
Ben Thompson from Stratechery asked a straightforward question about cancer risk and GLP-1s. He got kicked to a lesser model.
Someone asked about mitochondria. Same result.
J-Cal asked about fertilizer regulations live on the podcast to test it. Downgraded in real time.
Anthropic has since walked back the part about silently downgrading users for AI research. They now say they will disclose when they downgrade you. But they are still downgrading people. The surveillance is still running. The profile is still being built.
This is the company that once said it was against government surveillance.
They are now doing it themselves. To their own paying customers. For their own reasons. With no appeal process and no way to know it happened.
The developer community did not forget that.
WATCH THE FULL PODCAST ON @theallinpod
Like I've been saying for over 2 years now. AI today is great at regurgitating...
Ken Griffin was asked if AI could tell you which stocks to buy, before he bought the entire book of Leopold Aschenbrenner: "there are people that have a fantasy that LLMs are going to tell you which stocks to buy - and it's just a fantasy, just a fantasy"
"but wouldn't they be smart enough to tell you which stocks to buy?" - "no, because they're not smart, they're just programmed - they are very good at thoughtfully regurgitating"
this is him explaining why he thinks the productivity gains everyone is pricing in are still years away, why Nvidia reminds him of the California gold rush, and what he says is the one risk the market is completely missing
"it's not clear to me that we're going to get the productivity gains out of AI that the market's broadly hoping for - they're not about to happen in the next 12 or 24 or 36 months"
"this is like the gold mining in California - those that sold the picks made a lot of money - and Nvidia right now sells one of the most important ingredients in the AI story"
"the question with every investment is whether or not it's the right price - that's the fundamental of investing"
bookmark & watch the full conversation ↓
Stuart McClure retweeted
Ken Griffin was asked if AI could tell you which stocks to buy, before he bought the entire book of Leopold Aschenbrenner: "there are people that have a fantasy that LLMs are going to tell you which stocks to buy - and it's just a fantasy, just a fantasy"
"but wouldn't they be smart enough to tell you which stocks to buy?" - "no, because they're not smart, they're just programmed - they are very good at thoughtfully regurgitating"
this is him explaining why he thinks the productivity gains everyone is pricing in are still years away, why Nvidia reminds him of the California gold rush, and what he says is the one risk the market is completely missing
"it's not clear to me that we're going to get the productivity gains out of AI that the market's broadly hoping for - they're not about to happen in the next 12 or 24 or 36 months"
"this is like the gold mining in California - those that sold the picks made a lot of money - and Nvidia right now sells one of the most important ingredients in the AI story"
"the question with every investment is whether or not it's the right price - that's the fundamental of investing"
bookmark & watch the full conversation ↓
Gavin Baker: "I don't know another entrepreneur or another business that's a better bet on the future than SpaceX"
here's what Jensen Huang told them: a supercomputer like this normally takes three years to plan and another year to make it work - "we're talking about 19 days - Elon is an N of one"
this is them explaining why SpaceX went from not being an AI hyperscaler to number four in thirty days, what the return on the first Colossus cluster actually was, and why Grok put xAI on the Pareto frontier - one of only four companies that got there
"we know from Jensen that Elon brings data centers up faster than anyone - 122 days - speed is literally cost, because every day you're paying electricians and plumbers"
"Freda calculated a 55% IRR on Colossus one - if you can borrow money at six, seven, eight percent and invest in something with a 55% IRR, I'm not the most sophisticated thinker, but that math maths"
"Grok was on the Pareto frontier - the most intelligent 500 billion parameter model in the world - there are four companies on the frontier"
"in 30 days we went from not being an AI hyperscaler to being number four - and we passed a lot of companies, including Oracle"
bookmark & watch the full conversation ↓
Had an incredible discussion with James Shira the Global CIO of PwC about AI, cybersecurity and the future of humans at work. @WethosAI
youtu.be/4jHWv0CxXtA?is=Uf4q…
Most boards do not have an information problem. They have a decision operating-system problem.
The materials live in the board portal. The ideas and discussion lives in the meeting. The assumptions live in everyone's heads. Six months later, no one can reconstruct why the decision was made, or hold anyone accountable except the HiPO (highest paid officer).
This really needs to change... @WethosAI
Stuart McClure retweeted
McKinsey's newest decision-making research puts a hard number on something every executive already feels.
Senior leaders spend up to 70% of their working hours making decisions, and 57% of C-suite leaders say most of that time is used ineffectively. Only 41% of organizations report that their decisions actually align with their own strategy.
That's not a talent problem. Every one of those leadership teams has smart, experienced people in the room. It's a process problem: no reliable way to know, before the meeting, where the real disagreements are and which objections are worth taking seriously.
Fixing that doesn't require more meetings. It requires a clearer picture of the room before you walk into it.
#ExecutiveDecisionIntelligence #DecisionMaking #Leadership #CSuite #WethosAI
Source: mckinsey.com/capabilities/pe…
Boards have spent decades investing in finding the right CEO. They’re now realizing they need to invest just as much in helping that CEO make better decisions after they’re hired.
ft.com/content/7fc8f66e-b494…
The outlier is becoming the trend...
Boards aren’t granting CEOs time to “learn the business.” They expect judgment from the start, and the tolerance for ambiguity has collapsed. Successful leaders must arrive pre-oriented, understanding the real mandate, the hidden risks, and how decisions are made before they walk through the door.
fastcompany.com/91561607/boa…
Stuart McClure retweeted
The first experimental evidence of recursive self-improvement (RSI).
Autoresearching the autoresearch agent for eight days.
The result beats the harness we hand-tuned for two years, on held-out benchmarks: 🧵(1/7)
As always, an insightful and objective voice in the space. I have long been skeptical of any AI that will come close to human level thinking much less beyond it. But I have started to see the Gen3 AI intentions with Recursive Self-Improvement and Self-Learning and see the potential now. However, for any of these endeavors to work, you must understand how the human brain works and builds its patterns of thought and interaction. Without understanding the patterns, we will never achieve AGI.
AI is a prime objective for most CEOs but how many use it for decision intelligence and simulation to prevent crisis and catastrophe? Not many.
"This is what our [Bain & Co] 2026 CEO Agenda Survey reveals: not a crisis of leadership at the top but a gap between where strategy is set and how it is delivered."
Bain's 2026 CEO Agenda, published this week, found that fewer than half of surveyed CEOs believe their own organization is agile enough to execute the strategy it just set.
Not because the strategy is wrong. Because confidence fades somewhere between the decision and the execution of it, in the operating model, the speed, the follow-through. That gap doesn't show up in the strategy deck. It shows up three months later, in the version of the decision the organization actually implemented versus the one the leadership team thought it made.
Closing it starts with a simple question, asked before the commitment: does everyone in this room actually see the decision the same way?
#ExecutiveDecisionIntelligence #DecisionMaking #CorporateStrategy #Leadership #WethosAI
Source: bain.com/insights/the-2026-c…
"The threats are clear; the organizations aren’t ready
For most of the past century, strategy was a deliberate craft. The business landscape shifted slowly enough that leaders had time to see it coming, deliberate, and respond. Today, every decision a company makes is effectively a prediction, a bet on the future. Making better predictions about business shifts has become the first step in successfully navigating them.
Conviction turns those predictions into competitive advantage. When leaders have it, they can move faster and commit more fully. When they don't, they must spread their bets, keep options open, and watch industry signals more carefully."
Stuart McClure retweeted
In 2023, a CFO had 48 hours to decide whether to cut a product line generating $12M in revenue.
Most decisions at that level take three months. His team had the same data everyone has. What he had that most don't: a clear read on where his leadership team actually stood before the meeting started. He knew which objections were real and which were reflex. The decision got made in one session. No second-guessing. No revisiting.
What made the difference wasn't the data. It was decision readiness.
#ExecutiveDecisionIntelligence #DecisionMaking #CorporateStrategy #Leadership
Source: bain.com/insights/the-2026-c…
Bingo. The longer the context the less accurate the answers.
Prime Intellect engineer:
"everyone's bragging about a million-token context. here's what they don't tell you.
at 256k tokens GPT-5.5 scores 80% on retrieval. push it to a million and it drops to 36%. the model accepts the context, it just can't reason across it. people call it context rot."
in a 20-minute talk he explains why bigger context windows won't save your agents.
continual learning + training on your own traces + real environments - that's the fix.
Watch the talk, then save!
