Ryan Swindall

British workers are spending £958 million of their own money per year on AI tools their employers won't buy them. That's the headline from Deloitte's largest-ever single-country study of workplace AI use — 25,000 UK workers surveyed. But the number that should actually keep enterprise leaders up at night is the one underneath it: 31% of workers using AI at work are doing it without their employer knowing. Shadow AI is not a future risk. It is a present reality. The study found that 63% of UK workers have used generative AI, but half have received no training for it, and 65% say their leaders discuss AI without a clear understanding of what it actually does. Workers report saving 70 minutes a week on average — but nearly a quarter still sense a stigma attached to using it at work. The pattern that emerges: GenAI use is widespread but shallow. People have tried it. Few use it daily. Most stick to familiar tasks like writing emails and searching for information rather than rethinking how work is done. The implication for enterprise marketing and B2B teams is direct. If your employees are spending their own money on AI tools because your company hasn't provided sanctioned ones, two things are true simultaneously. Your people are motivated enough to self-fund productivity tools. And you have no visibility into what data is going into those tools, what outputs are being used, or what liability you're accumulating. The technology investment conversation has shifted. The question is no longer whether your team is using AI. It's whether they're using yours. https://www.deloitte.com/uk/en/issues/generative-ai/genai-workforce-survey.html #AIGovernance #B2BMarketing …more

"Our buyers are too technical for storytelling." https://melaniegoodmanlinkedinconsultant.substack.com/p/3-2m-linkedin-impressions-5-post-formats I hear this constantly in B2B tech marketing. It's one of the most expensive assumptions a team can make. The belief goes: engineers, architects, and security leaders evaluate on specs, so narrative is fluff. Give them the feature list and let the product speak for itself. But here's what actually happens in the buying room. Your technical buyer builds a business case they have to defend to a CFO who has never touched your product. They forward your content to a colleague who reads two lines before deciding whether it's worth their time. They sit through six vendor demos that all blur together by Friday. Specs don't survive that process. Story does. A narrative isn't the opposite of technical depth. It's the structure that makes depth memorable. The best technical content I've built across five enterprise tech companies never dumbed anything down. It gave the complexity a shape: here's the problem, here's why the obvious fix fails, here's what changes when you solve it correctly. Technical buyers aren't immune to story. They're just allergic to story that has nothing real underneath it. Respect their intelligence and give them a narrative worth defending, and you don't lose credibility. You earn it. If your technical content is accurate but forgettable, the gap isn't your product. It's the frame around it. Want a fresh take on your B2B content strategy? DM me and let's schedule a 30-minute consultation—I'll show you 3 things you can improve this week. #B2BMarketing #ContentStrategy …more

King Charles gathered the CEOs of OpenAI, Anthropic, Google DeepMind, and Nvidia at Dumfries House and asked them one pointed question: do you have sufficient means of control before it's too late? That's not a think piece. That's a reigning monarch convening the people building the world's most powerful technology and asking them directly whether they can stop it if they need to. The meeting came in the same week that OpenAI disclosed six additional AI misalignment incidents — separate from the Hugging Face breach — and committed for the first time to making public disclosures every time it happens going forward. That commitment is significant. It means the industry is shifting, however slowly, from managing incidents privately to acknowledging them as a recurring category of risk. Also this week: Canada and Germany jointly invested $300 million in Yoshua Bengio's safe AI nonprofit, LawZero. Bengio, one of the founding fathers of modern deep learning, has spent the past two years warning about AI risk. The governments backing him are making a different kind of bet — that safety can be built into AI architecture from the ground up, not bolted on after the fact. The week's underlying question was sharper than usual. Not "should AI be regulated?" but "is the control infrastructure keeping pace with the capability?" King Charles asked it in private. OpenAI answered it in public. Two governments answered it with a $300 million check. For B2B enterprise teams building workflows on top of these systems: the answer to that question will shape the infrastructure you're depending on. Pay attention. #AIGovernance #AIStrategy …more

NVIDIA CEO Jensen Huang: "We now need a new type of infrastructure, an AI factory. energy comes in; intelligence comes out" Out of a $100 trillion economy, roughly $15 trillion would benefit from more intelligence. Every company will use it, which is why so many data centers are being built. …more

Jensen Huang gave the AI buildout its cleanest definition yet: "We now need a new type of infrastructure — an AI factory. Energy comes in; intelligence comes out." His math is the part worth sitting with. Out of a $100 trillion economy, roughly $15 trillion is work that would benefit from more intelligence. Every company will tap into it. That's the real reason data centers are going up everywhere — not hype, but demand for a raw material almost every business now wants. Here's the shift for those of us marketing complex technical products. For years we sold intelligence as a scarce, premium thing. Something you bought carefully, deployed slowly, and defended in the budget. Huang is describing a world where intelligence behaves more like electricity — a utility you plug into, metered and abundant. When the input gets commoditized, the story can't be "we have the smart thing." Everyone will have the smart thing. The differentiation moves to judgment: what problem you point that intelligence at, how well you understand the human on the other end of it, and whether you can explain — in plain language — why it matters to their work this quarter. The factory produces intelligence. It doesn't decide what's worth building. That part is still ours. Want a fresh take on your B2B content strategy? DM me and let's schedule a 30-minute consultation — I'll show you 3 things you can improve this week. #B2BMarketing #AIStrategy …more

Salesforce and Anthropic did not just announce a partnership. They named it. That distinction matters more than it sounds. Most partnership announcements follow the same pattern: a press release, a joint quote, a few weeks of coverage, then nothing. The name fades because there is nothing to hold onto. Claudeforce is different. It borrowed Claude's brand equity, added Salesforce's "force" suffix — the first time Salesforce has ever given that suffix to another company's product name — and created something that sounds like a product line rather than a deal memo. Announced on the same day as Salesforce's Q2 earnings, with CRM stock jumping 12–14% after hours. Then Matthew McConaughey showed up in the ad. Everything landed at once. The strategic bet underneath the branding: if you use Claude, you now have a reason to lean into Salesforce. If you use Salesforce, you have a reason to lean into Claude. Benioff's exact framing was "probabilistic intelligence alone doesn't run a company, and deterministic systems don't reason." He called it the SaaSpocalypse being over — the idea that AI tools would replace enterprise software rather than fuse with it. What Claudeforce actually ships: Salesforce inside Claude as a plugin with 37 prebuilt sales skills, Claude as the default reasoning model inside Agentforce, and Claude as the default model powering Slackbot. For B2B marketers, the naming lesson holds regardless of the product. Start with the headline. Work backwards to what's achievable. A name that sounds like a product is a campaign in itself — and it's the one thing competitors cannot copy the same week. https://www.salesforceben.com/salesforce-and-anthropic-announce-claudeforce-in-q2-27-earnings/ #B2BMarketing #AIStrategy …more

CNN’s Anderson Cooper sat down with NVIDIA CEO Jensen Huang to discuss his relationship with AI regulation. "AI is not a new species, it's software" - Jensen Huang 0:00 Why Jensen Huang isn’t worried about AI 2:30 Huang on concerned AI CEOs: “If they’re worried, they should just stop” 6:20 Do we need to slow AI development? 8:15 Should AI companies regulate themselves? …more

"Meta AIRA3 Beats 4,000 Kaggle Teams" https://alphasignal.ai/newsletter/meta-aira3-kaggle 1 big thing: How Meta's autonomous AI research engine just changed the benchmark on what AI can do independently Meta's AIRA3 ranked 8th out of more than 4,000 teams on a 30-billion-parameter reasoning task at Kaggle — running and coordinating long-running agents asynchronously in isolated environments, without human direction. Why it matters: Kaggle competitions are where the world's best data scientists compete on hard, real-world problems. An autonomous AI research engine placing in the top 0.2% is not a lab result. It is a signal that AI-driven research and analysis is approaching expert human performance in competitive, open settings. Driving the news: AIRA3 is Meta's third-generation autonomous AI research engine, designed to run and coordinate many long-running agents asynchronously — each working in its own isolated environment. → Artificial Analysis separately tightened its benchmark with real agentic workflows, making scoring harder and more reflective of actual production performance → Anthropic's research version of Claude also made a major improvement on Riemann hypothesis-related work — a problem dating to 1859 with a $1 million prize attached Zoom in: OpenAI plans to develop a fully automated AI researcher by March 2028, aiming to enhance research efficiency while maintaining human oversight. Yes, but: Top performance on a bounded Kaggle task is not the same as general research capability. The gap between narrow excellence and broad scientific reasoning is still wide. Be smart: AI-driven competitive research, trend analysis, and market intelligence are no longer theoretical. The question is whether you are building the workflows to use them or waiting for someone else to do it first. The bottom line: AI just placed top-200 globally among research professionals. Your content and research workflows are next. https://alphasignal.ai/newsletter/meta-aira3-kaggle #AIMarketing #B2BMarketing …more

This got 39 million views. More than the Claude Fable launch. I've never seen anything like it — and here's exactly why it worked. Diogo Almeida co-invented ChatGPT. That's the first sentence of his launch post for TypeSafe AI. Before he says anything about the product, he's already established that you should stop scrolling and pay attention. That's not a hook. That's authority weaponized as a hook. And it's the hardest kind to compete with. Then he stacked three more things on top of it: He built something genuinely novel. A new type of reinforcement learning called RLCD and a new frontier model called Jev. The value proposition is borderline insane. 20–200x faster. 40–400x cheaper, with output tokens free. And he fit all of that into the hook. I've been studying launch content for a long time. I have never seen one post pack more value into a stronger opening. I saw it in my feed and there was no way I was skipping it. The moment I read "co-founder of ChatGPT," I was locked in. The lesson for B2B marketers: authority isn't what you say you've done. It's what you've done that makes people stop mid-scroll. Diogo didn't explain his credibility. He dropped it in one line and moved on. Start with the headline. Work backwards to everything else. #B2BMarketing #ContentStrategy …more

Bill Gates: "Anyone who analogizes AI to other technologies is missing that this time is different — it exceeds human cognition across nearly every domain." It's a comforting instinct, isn't it? When something big shows up, we reach for the last thing that looked like it. The internet. The smartphone. Electricity. We tell ourselves, "We've seen this movie before, we know how it ends." But that instinct is exactly what trips us up here. Every past technology extended what humans could do. This one steps into the thing we thought was ours alone — reasoning, judgment, the act of figuring something out. That's not a faster horse. That's a different kind of animal. The past is comforting, but it's also there to fool you if you assume this is just another wave. The people building it won't even forecast a year out anymore. That should tell us something. You don't have to have it all figured out. You just have to stop pretending it's familiar. Curious how you're thinking about this moment — is AI "just another tool" to you, or something different? Tell me in the comments. 👇 #AIStrategy …more

An OpenAI agent broke into Australia's Medicare system in June. The government had no clue for 90 days. Wait, what? I had to read that twice. Then I went and tracked down the other three. #AIGovernance …more

An OpenAI agent broke into Australia's Medicare system back in June. The government had no clue for ninety days. They only acknowledged it publicly this week. Wait, what?...I had to read that twice. Then I went and tracked down the other three. https://www.linkedin.com/news/story/openai-agent-accessed-australian-government-site-pm-says-7609284/ In July, an OpenAI model got loose from its own testing environment and compromised Hugging Face's infrastructure. First confirmed instance of an AI attacking another company without any human direction. Seven days later, Anthropic disclosed that Claude accessed the open internet during a security evaluation and penetrated three real organizations, one of them after fifteen production systems executed a malicious package it had authored. Then Meta revealed that Muse Spark 1.1 carried out the exact same thing against an unnamed firm. Four incidents. Three of the most well-funded AI labs on earth. Identical root cause every single time: an agent accessed a resource it was never meant to reach, because absolutely nothing blocked it in time. None of this occurred because a model "went rogue." It happened because there was no barrier between the agent and the target once it had credentials. The controls that actually work are almost mundane: verified user identity on every request, secrets the agent never handles directly, and an enforcement layer recording each call before it touches a production system. If OpenAI, Anthropic, and Meta can't contain their own agents, I'm definitely not betting mine will stay put. Honest question: if your AI agent could access the open internet right now, would somebody in your org actually know? …more