Databricks announced Vibe Data Modeling, a multi-model LLM agent that enables you to describe your business and create a data model that ships in a single notebook and can become a fully deployed model in the Unity Catalog. According to Databricks, a minimum viable model can be deployed in less than two hours.
Salesforce Ben argues that Salesforce's penchant for renaming every product repeatedly hampers the company. What used to be a running joke about renameforce.com is confusing customers. Just imagine what'll happen when Salesforce eventually becomes Agentforce.
People have the AI bubble all wrong. It's not a technology bubble. It's a real estate bubble that rhymes more with the subprime mortgage crisis in 2008 than the dot-com boom. That's the takeaway from a post by Groundbreaker.
The gist:
"The market is pricing AI as a technology cycle when its actual anatomy is that of a credit-driven real estate cycle - which is precisely why the 2008 mechanics apply - and the two break for entirely different reasons.
Technology cycles are driven by innovation and adoption; their risks are obsolescence and competition; they live or die on whether the product is wanted, and they can de-rate slowly as the future is repriced.
Real estate cycles are mechanical: leverage, hard assets, occupancy - debt-financed construction at scale, commercial leases disguised as take-or-pay contracts, and long construction lags that guarantee supply arrives after demand has turned. Walk down the AI build-out and every feature is a property development in disguise: a data center on entitled land, financed with debt against the structure and leased to tenants on take-or-pay terms."
Hat tip to Wall Street money manager George Noble for flagging the Groundbreaker post.
Anthropic posted interesting research on J-space, a set of neutral patterns in Claude that reveals internal thoughts. J-space wasn't programmed, but has emerged.
"We present evidence that a similar distinction has emerged in modern language models like Claude. We find that Claude has developed a small collection of internal neural patterns that, compared to all its other internal processing, play a special role...
We call the collection of these patterns the J-space—named after the technique we used to find them, involving a mathematical concept called the Jacobian. Each J-space pattern is linked to a particular word. But when one of these patterns lights up, it doesn’t mean the model is saying that word—just that the word is on its mind. If you've heard of language models having a "scratchpad" or “chain of thought”—text they write to themselves while reasoning—the J-space is something different. It operates silently, in the model’s internal neural activations, allowing the model to think about a concept without writing it down. Notably, the J-space wasn’t designed or programmed by us, but instead emerged on its own during Claude’s training process."
Netflix is reportedly studying data to see why viewers quit shows after the first season. Some shows see a 30% to 70% drop. I'll save Netflix some time on the data crunching.
1. There aren't a lot of shows that necessitate a long-term investment. Netflix is way mediocre.
2. The CX is terrible and if you can avoid Netflix you will.
3. Why is the CX so terrible? It's one thing to have to tell your credit card company when you're traveling. It's another to tell Netflix. The hurdles, CX loops and general pain with explaining how you may have a second home that shouldn't require a second subscription is getting to be a bit much.
4. Netflix raises its prices regularly so expectations have gone past the algorithmic slop the streaming provider offers.
So go ahead and study the data, the but the short answer for Netflix is it kinda sucks.
Cybersecurity vendor Sysdig said it has documented the first case of an agentic AI ransomware. In a blog post, Sysdig documented JADEPUFFER, which carries out an ransomware operation end-to-end via an LLM. Typically, ransomware is run by a human. JADEPUFFER is an agentic threat actor that can prioritize targets and adapt in real time.
Microsoft said it is launching the Microsoft Frontier Company, a unit that will deliver AI transformation. The company said it is investing $2.5 billion in the new unit and embedding 6,000 industry and engineering experts.
In a statement, Microsoft said:
"This goes beyond what has been labeled as Forward Deployed Engineering (FDE) and will be the largest, most capable, outcome-driven engineering organization in the industry."
Microsoft's move comes after AWS launched a forward deployed engineering effort.
According to the Financial Times, OpenAI has discussed giving a 5% stake to the US government. The early talks are aimed at clearing political hurdles as it releases models. Reaction is mixed to say the least.Say what you will about OpenAI CEO Sam Altman, but he's a master of raising money and playing the game.
Amazon said its Amazon Leo unit has completed 14 missions and launched more than 375 satellites. The company started scaling its satellite deployment in April and is not the third-largest constellation in orbit.
Meta launched a $20 month Meta One Premium subscription for its glasses. Free users (nevermind that you paid for the creepy Ray Bans in the first place) will have limits on conversations. I'm sure this'll go over well.
"No subscription is required to use AI glasses. All AI glasses owners get free monthly usage for certain features. If you reach your free monthly usage limit, you can upgrade to a paid Meta One Premium plan, or wait until your free limit refreshes each calendar month.
If you reach your paid subscription usage limit, it will refresh at the start of the next billing cycle.
Conversation focus is currently available for free for 3 hours per month, or for 15 hours per month for Meta One Premium subscribers."
Here's what's next: If you don't pay up, Meta will limit your ability to see out of the glasses. Eyesight is now under a consumption plan.
Yay, Anthropic's latest models are back after a lot of thrash with the US government. The biggest takeaway from Anthropic's statement and my recap is this:
The same jailbreak that sidelined Fable 5. Anthropic said Amazon researchers found a way to bypass Fable 5's safeguards to identify software vulnerabilities. Anthropic said its testing found that other models had the same ability including Claude Haiku 4.5, Sonnet 4.6, Opus 4.6, Opus 4.7, Opus 4.8, GPT-5.4, GPT-5.5 and Kimi K2.7.
Schneider Electric acquired Cognite for $3.1 billion in a deal that'll beef up its industrial AI efforts. The plan is to combine Cognite with Scheider Electric's AVEVA industrial software unit.
