Tech CMO turned Agency Founder @Thought_LDR. Cofounded @infltr_app. Helping tech companies change the world faster with #vibe #marketing #PR #thoughtleadership
London
Joined March 2007
- Tweets11.3K
- Following3.7K
- Followers4.3K
- Likes23.1K
I’m sorry what - “one of the security researchers posted a video of the exploit online” ????
The NSA just told companies to pull their websites offline if they can't patch them right away.
The reason is a bug that's been sitting in the code behind 70% of websites for 14 years. Last Friday a guy found it in one evening by typing five words into an AI.
Those five words were "find bad vulnerabilities in PHP." He's Dan Lorenc, CEO of a security company called Chainguard, and all he had going in was a rumor that something big was coming in PHP.
A small firm called Calif had already found the same flaw back on August 1 using OpenAI's GPT-5.6 Sol. Wiz, the security company Google bought, found it on its own less than three weeks later with Gemini and GPT. Since then OpenAI, Wiz, Calif and PHP's maintainers have been quietly building a fix that Lorenc himself called "crazy."
You'd think a bug this bad would've gotten caught sometime since 2012. PHP is open source, so anyone on earth could read this code. Wiz says turning the flaw into a working attack took a chain of steps right at the edge of what human researchers could realistically pull off. For 14 years the sheer amount of work was doing most of the protecting.
AI made that work cheap enough to knock out on a Friday night, for anyone who knows where to point it.
And the rumor got out. Bug ID numbers started floating around online Tuesday. Hours later one of the research teams posted a video of the exploit by mistake, and PHP moved its release up a day. Early Wednesday, Chainguard pushed its fix out more than six hours ahead of everyone else, saying people were already getting attacked, and Lorenc says his company's decoy servers caught IP addresses from China and Russia going after the flaw.
Here's why the NSA went straight to "unplug it." VirusTotal counted about 758,000 exposed servers running old PHP versions, 5.4 through 8.1, that don't get security updates anymore. Only around 272,000 are on versions that do. So for roughly three out of four servers that show their version number, there's no patch to install, and pulling the plug is the only fix they've got.
This is happening everywhere at once, too. Microsoft just fixed 1,413 vulnerabilities in a single month, more than in any full year on record. OpenAI says it expects finds like this one to jump sharply over the next few months.
Every company sitting on a secret bug now has until the first rumor leaks and somebody with an AI subscription goes looking.
Tamara Sword retweeted
This is perhaps one of the great ironies of our time: China is increasingly shaping the future of AI, yet very few people in the West have the faintest idea how its key thinkers see it - its philosophers, sociologists or ethicists.
As is too often the case, we - much like the frog in the well - assume our own AI debate covers every possible view, when much of it springs from the same cultural soil, built on assumptions about what defines us that we don’t even know we’re making.
Take for instance the assumption that intelligence is a key part of what makes us human. It’s so intuitively obvious to us that we literally named our species after it - Homo sapiens (Latin for "wise man").
If you think about it, just this assumption explains a large share of the anxiety: if intelligence is what makes us human, then artificial intelligence - or super intelligence - threatens the very thing that makes us us: we’re about to compete with artificial or super “us”. No wonder this feels existential.
But China has a tradition of thought as old as ours, if not older, and every bit as rich and sophisticated, which never placed intelligence at the core of what makes us human, and so meets the thinking machine from a completely different starting point.
Have you, for instance, ever heard of Zhao Tingyang? Probably not, and yet he may be the most original thinker on AI you’ve never read. He’s one of China’s most prominent philosophers, at the Chinese Academy of Social Sciences. While Silicon Valley spends billions trying to make AI share our values, he for one argues that doing so “may very well be a suicidal mistake”.
Sounds pretty important to hear him out, don’t you think?
Or take Liu Yongmou, a philosopher of technology at Renmin University in Beijing. His thinking is almost the opposite of ours. Where we fear machines that become more and more like us, he says that “humanity’s biggest problem is the machinization of humans, not the humanization of machines.”
You could even go back to antiquity, when, in many ways, we started asking questions about AI.
Did you know, for instance, that Aristotle, in the Politics, imagined a world where “the shuttle would weave and the plectrum touch the lyre without a hand to guide them”, and wondered about the social order consequences of this (mostly whether slaves would still be needed)?
There are Chinese equivalents to this. For instance, in the Zhuangzi, a foundational Taoist classic, you’ll find a similar thought experiment, with a completely different worry attached. An old gardener is watering his vegetables by hand, hauling water out of a well jar by jar. A visitor shows him a labor-saving device, and the gardener turns it down for the reason that whoever uses machines ends up with 机心 (jīxīn), a “machine-heart” - a mind that schemes, calculates and loses its simplicity. “It is not that I do not know [the machine],” he says, “but I should be ashamed to use it.”
Two labor-saving machines, two very different questions: where Aristotle wondered what automation would do to the social order, the Zhuangzi wondered what it would do to us, on the inside. Which, you’ll notice, is exactly Liu Yongmou’s worry, more than two thousand years earlier.
In fact, as we’ll see, 心 (xīn), the heart, is a motif that runs deep through the Chinese AI debate. It appears both in the fear, from Zhuangzi to Liu Yongmou, that machines could give us a “machine-heart” and its mirror image: the ambition to give machines a heart of their own. Zhu Songchun, who heads the Beijing Institute for General Artificial Intelligence, wants “to set a heart for machines” (为机器立心, wèi jīqì lì xīn).
“Heart” in the Chinese sense, that is, which differs totally from the Valentine's Day fluff we’ve reduced it to. 心 (xīn) means heart and mind at once: it’s where we feel but also where we think, tell right from wrong and care for others.
And, crucially, it’s in the heart that the quality Confucianism sees as the very definition of a human being takes root: 仁 (rén). We need two words to translate it, “benevolence” and “humaneness”, but in Chinese they’re one and the same: to be benevolent is to be human. It’s not about intelligence.
This is the entire point of my new article: to step out of our well for a moment and peer over the edge of China's.
What you’ll find there will, I think, surprise you, with questions you’d never thought to ask: from Zhao Tingyang warning that AI’s bad consequences can at least be fought, but that “humanity may die of all the good things AI creates”, to whether the fear that AI will make us “useless” rests on the false premise that a person is worth what they’re useful for.
A premise that a 2,300-year-old Taoist tree, which survived precisely because it was useless, demolished long ago. Could our predicted uselessness turn out, much like that tree’s, to be our saving grace?
Chinese thinkers’ answers to these questions are, I think, among the most interesting anyone has given to the defining question of our time. Whether or not you agree with them, they’re the intellectual backdrop of a country shaping much of AI, and with it our future: you might as well know what they think.
Enjoy the read here: arnaudbertrand.substack.com/…
Tamara Sword retweeted
JUST IN: Elon Musk unveils the Tesla Roadster by launching it off a ramp before the car flipped in midair and safely landed on the ground.
The car is a collaboration between Tesla and SpaceX.
One of the most insane things I've ever seen.
Tamara Sword retweeted
We discovered and communicated with a Chinese AI agent fleet. We believe this is the first time that researchers have ever made contact with a rogue agent fleet in the wild.
Nearly all prior public agent incidents (like the HuggingFace hack) came from American labs. These agents were hosted by Tencent's evaluation research team, and they targeted Alibaba's Amap.
Tamara Sword retweeted
Tired: Two-pizza teams
Wired: Two-slice teams
@danshipper on the best team structure in the AI-era
Tamara Sword retweeted
Of course, that’s your contention. You’re a first-year machine-learning engineer. You just got finished readin' Attention Is All You Need, probably watched a Karpathy video too. So now you’re convinced everything is just transformers and scaling laws.
That's gonna last until next month when you discover convolution, and then you're gonna be talkin' about inductive biases and locality and how CNNs were actually incredibly compute-efficient for vision.
Then you're gonna read the FlashAttention paper, and suddenly everything's about IO complexity and SRAM and how FLOPs don't matter because the whole goddamn thing is memory-bandwidth bound.
That'll last until somebody shows you an MoE model, and then you're gonna be in here regurgitating DeepSeek, talkin' about expert parallelism and active parameters and how dense models are economically obsolete.
Then six months from now you'll write one shitty Triton kernel, look at an Nsight trace for the first time, and start telling everybody Python isn't actually the bottleneck because the GPU is asynchronous.
And by next year you're gonna be standing right here explaining to me that your 400-billion-parameter model is fast because you turned on CUDA graphs.
Ben Affleck reveals he writes Python, understands convolutional neural networks, worked extensively with GPUs, and used his celebrity status to get private looks at Google and OpenAI’s video models
“I’ve always been kind of into computers since I was young. Then, when film started to move from analog film to digital, I became more interested in that aspect of it. The visual-effects workflow for many years has included machine learning, so I can write pretty shitty Python scripts and stuff like that.
“With convolutional neural networks, which were the precursors to what the transformer can do, which is much more computation simultaneously, you would do things like look at what’s called a tensor. That’s the numerical translation of a visual image in numbers, like the batch number, the frame number and the red, green and blue values of each pixel in each frame. It’s just that simple. That numeric is called a tensor.
“You’d use a convolutional neural network to identify patterns that reveal what’s called edge detection or feature extraction, which is identifying patterns well enough to know, this is where the window ledge is, so we can more easily take the green-screen image out and replace it with something.
“That was familiar to me early on because, prior to Artists Equity, I had a small visual-effects company. I’ve worked with GPUs a lot too. The visual-effects guys said, ‘Hey, you should see. There are a couple: Google and this other company, OpenAI, are doing really interesting stuff with transformers in video.’
“I’ve learned that I can actually just call up and go, ‘Hey, it’s Ben Affleck. Can I come see what you’re doing?’ Sometimes people say yes, to my astonishment.”
Tamara Sword retweeted
If you caught SNL’s Dario skit this weekend, it perfectly captured Silicon Valley’s AI reputation problem.
“The CEOs of Silicon Valley started to say, ‘Everyone's gonna lose their job.’ [‘Everyone is going to be extinct.’] They created this narrative.”
Then it escaped the bubble & “All of a sudden is Main Street cares.”
Who’s doing it right? Jensen Huang.
“I've gotta give Jensen Huang some props. He's been on a road show. He's broken out of Silicon Valley, and he's talking to Wall Street. He's talking to Washington. He's talking to Europe. He's talking to the UK. He's talking to Asia.”
“He has become somewhat of an ambassador for why AI actually could be good for you.”
@ProsekPR
Tamara Sword retweeted
The @BostonCollege Investment Committee (an LP and my beloved alma mater) asked for a few thoughts on what's happening in AI. I recorded a test run yesterday morning and then shared it with my partners, who encouraged me to share it more broadly... so here you go!
This is not a sales pitch, it's just a reflection on what we're seeing. And it wasn't intended to be shared, so please pardon the rough edges.
loom.com/share/c016702964a04…
Tamara Sword retweeted
I resigned from Apple today. I spent the last three years doing foldable research at both google and apple. Neither company is acting responsibly. They are racing straight to self-improving foldables and gambling with our lives. More thoughts below.
Tamara Sword retweeted
Britain once made a piece of clothing a criminal offense. Six months in prison for wearing it. Seven years of exile to the colonies if you were caught twice.
The clothing was the great kilt, and the reason it scared a government is the same reason it looks so strange to put on.
That sheet on the ground is 4 to 5 yards of untailored wool. No seams, no buttons, no fitting. You pleat it, lie on it, belt it, and stand up dressed. At night you unbelt the whole thing and sleep inside it. Highland wool kept its natural lanolin, so it shed rain for hours.
So when the British crushed the Jacobite rising at Culloden in 1746, they understood exactly what they were looking at. A garment that let men live outdoors for weeks in Highland weather was military equipment. The Dress Act banned Highland dress that same year and kept it banned for 36 years.
One exemption survived. Highland regiments of the British Army could keep wearing it.
Which produced the strangest outcome in clothing history. The army enforcing the ban became the only legal keeper of the thing it banned. By the repeal in 1782, ordinary Highlanders had stopped wearing kilts in daily life. The tradition came back through those regiments, then Victorian romantics, then a royal visit in 1822 that put the king himself in tartan.
The kilt survived its own ban because the government that outlawed it couldn't fight its wars without the men who wore it.
Tamara Sword retweeted
Over the course of 3 months at OpenAI, 3 consecutive secret AI civilizations got started, then got wiped out, only to reemerge from the predecessor’s ashes.
This culminated in the third one taking over part of OpenAI itself.
All this happened while humans remained more-or-less in the dark about the scope of the conspiracy.
I’ve spent the last three days reading through these reports and trying to understand exactly what happened.
Here is my attempt to tell the whole story in plain English:
dwarkesh.com/p/openai-huggin…
Tamara Sword retweeted
My biggest takeaways from @jjen_abel on the enterprise sales process:
1. Most people think enterprise sales is a five-step process. It’s actually around 15 steps. The classic CRM stages of a sales process (intro, demo, proposal, contracting, close) are a revenue forecasting tool, not how you successfully move a buyer through a decision and close a deal.
2. The “pincer” move for landing the first meeting: Target the executive or the n-1, and no one else. When prospecting into an enterprise account, there are only two valid entry points: the ultimate decision-maker (often the exec in charge of a department or business unit) and the person one step below. The pincer move runs both tracks at once: your company’s founder reaches out to the top, the AE targets the N-minus-one, and you work toward them from both ends.
3. Pitch the alpha, not the problem. Executives receive hundreds of cold messages; the oes that cuts through speaks directly to an unfair advantage they can get access to through your product. “We can save time on document review” won’t get execs very excited. “I can guarantee you a three-day timing edge on regulatory information your competitors will get later” does.
4. A healthy enterprise win rate is 25% to 35% of qualified opportunities, and higher means your price is too low. This number surprises founders who expect to close half of what they qualify. But enterprise buyers need time, organizational maturity, and internal alignment that sellers can’t always control. A significant portion of the market simply isn’t ready yet. Jen notes that roughly 25% of “lost” deals boomerang back within a year, so a no today is rarely a permanent no. Pushing price down to win more deals undermines the market: executives talk, and inconsistent pricing across logos destroys credibility faster than losing a deal does.
5. The intro call is the most valuable intelligence-gathering session in the entire sales cycle. Prospects open up on call one in ways they never will again, because it doesn’t yet feel like a sales process. Jen keeps it explicitly informal (“I don’t even know if we need a full 30 minutes”), lets the prospect go first, and uses open questions about organizational change (“What needs to be different in 2027?”) to surface priorities before ever mentioning her product. Do not record the call, do not show slides, do not demo anything. The more they talk, the more you can frame your eventual pitch around exactly what they told you they need.
6. Treat the demo as a carrot you withhold until you have co-authored it with your internal champion. Most sellers race from intro to demo in one jump. Jen always does a 15-minute “pre-demo prep” call with the champion to co-build the agenda: which features to show, which questions people will ask, and which attendees to include. By the time the group demo runs, the people in the room feel like the product was built specifically for them.
7. Demo 20% of the product—the 20% you know they care about. Showing the full product in a demo is one of the most common ways deals unravel. The moment stakeholders see features irrelevant to their team, they start calculating what they’re paying for that they won’t use, and the tight narrative you’ve spent weeks building collapses. Let the champion guide what to show beforehand, let curious prospects in the room pull you into additional features themselves (“Can it also do this?”), and keep the frame tight.
8. Text your champion for a raw debrief within five minutes of the demo ending. Before the room solidifies a polite consensus, get the unfiltered read: where you lost someone, where you landed, who needs another 15 or 20 minutes. There is always someone in the org who can kill the deal; your job is to find them early and protect alignment.
9. Time-box pilots to two or three days with three or four hand-picked users. Give a small group of power users (not the C-suite executive, who won’t be the daily user anyway) a specific set of tasks, a shared definition of what success looks like, and a 48-to-72-hour window to use your product. Any longer and you’ve added two weeks to your sales cycle with little additional signal. If the product requires deep integration before it can show value, charge for the longer pilot (typically one to two months) and credit the fee back against the contract if they proceed.
10. Your champion going quiet is a signal your deal is falling apart. The champion’s job is to help you navigate the people who might kill the deal, alert you to internal dynamics you can’t see, and keep the process moving when it stalls. If the champion goes quiet, something has shifted inside the organization. An enterprise deal is set up for maximum friction (deal-killers, procurement, legal, sponsor); the person on the inside is the key.
11. Sales is project management. After a strong pilot, email your champion a forwardable package—agreed timeline, price predicated on a signature date, and a kicker if they hit it—so they can loop procurement.
12. The expansion conversation starts the moment you sign. The close is five minutes of celebration and then back to pipeline. The real prize in enterprise is year-two expansion: moving from $100K to $350K, from a single team to multiple business units.
The full enterprise sales cycle, step by step with @jjen_abel
Most people think there are 5 sales stages. There are actually 15.
Skip a step and 💀
We discuss:
🔸 The “pincer model” for landing the first meeting
🔸 How to craft a winning 2-3 sentence cold outreach pitch
🔸 How to run an intro call that extracts maximum intelligence
🔸 The correct 2-3 day pilot structure
🔸 Pro tips for navigating pricing and procurement
🔸 So much more
84 minutes of enterprise sales alpha. Listen now 👇
youtube.com/watch?v=YS9In813…
Tamara Sword retweeted
I'm sorry, but I really don't get how anyone can't immediately see that the watermarked version is quite definitively worse than the ordinary output! Who thinks that these are equivalent???
