Digital twins are virtual models of physical entities used to optimize business operations. Success lies in understanding best practices, defining business goals, testing via prototyping, assessing organizational readiness, and adopting an agile approach. RT @antgrasso

Key findings about how Americans view #AI buff.ly/QLpihPY @pewresearch Cc @terence_mills @LaurentAlaus @Ym78200 @Nicochan33 @Analytics_699 @Fabriziobustama

COOs, are your efforts delivering the results you expected? Supply chain transformation activity is at a four-year high, yet only 10% of companies met all three of their top priority targets last year. It's time to orchestrate the value chain as one interconnected system: bit.ly/4xPLzSb

The Private Equity Due Diligence Process: #AI vs. Traditional Approach buff.ly/OmrSELF @V7Labs Cc @EvanKirstel @SpirosMargaris @andresvilarino @ahier @enilev @Khulood_Almani @bamitav @FelipefajardoP @bimedotcom @andresvilarino

Controlling Reasoning Effort in LLMs buff.ly/8dhuAdR @rasbt #AI #GenerativeAI Cc @helene_wpli @floriansemle @jblefevre60 @pierrepinna @Ym78200

Findings show that a structured co-adaptation process can be conducted in LMIC settings and can generate locally relevant adaptations while maintaining core intervention functions. Link: gh.bmj.com/content/11/9/e024… #mentalhealth #cambodia #vietnam

AI can reveal patterns in clinical data, but a care plan still needs the doctor who understands the patient and the full medical context. A model may notice that a patient’s test results resemble cases that responded well to a certain treatment. rt @antgrasso.

Global standards define market entry long before products compete, setting technical conditions that shape certification and interoperability. Companies that align engineering with compliance reduce redesign cycles and enter markets with fewer barriers. Rt @antgrasso.

Polished AI answers are not proof of truth. Weak data can still produce confident language with very little substance behind it. So where should organizations look first? At the information behind the answer. Microblog by @antgrasso about #AI #DataQuality #ResponsibleAI Generative AI estimates which response is most plausible from learned patterns and the context it receives. That process can produce fluent language even when the company data underneath is incomplete or outdated. A sales assistant, for example, may confidently confirm that a product feature is available in a certain market. If the product records are wrong, the answer will simply express the error more convincingly. Better prompts can help the system understand the question. They cannot repair missing facts. Organizations need reliable business records and clear ownership of their accuracy. AI outputs should then be tested against real use cases before they reach customers or influence a decision.

Digital technology expands what healthcare can do, but to make the most of that potential, different systems need to understand each other. Adding new tools is not enough if their capabilities remain isolated across the healthcare system in daily care. Microblog by @antgrasso Think of a lab result moving from one healthcare system to another. Sending the number is not enough. The receiving system also needs to know what the result refers to, which unit was used, when it was recorded, and which patient it belongs to. If that context is missing or interpreted differently, the data has travelled, but part of its meaning has not. This is what interoperability helps solve: different systems can exchange information and interpret it consistently. Standards provide a common language. HL7 FHIR (Fast Healthcare Interoperability Resources), developed by Health Level Seven International, defines ways to structure and exchange health information. DICOM (Digital Imaging and Communications in Medicine) does the same for medical images and related data. The same principle applies when information moves between clinical systems and connected devices. Data should not need to be manually “translated” every time it crosses a boundary. And understanding each other does not mean giving everyone access to everything. Privacy and authorization still determine who can see and use the information. Good interoperability lets health data move without leaving its meaning behind. #DigitalHealth #Interoperability

Slamming the door on #AI based #cybersecurity threats buff.ly/XILAIqa @InformationWeek Cc @Ym78200 @akwyz @LaurentAlaus

Despite its complex name, a decision coherence system has a simple goal: helping different teams make choices from the same business context. AI brings shared information and visible trade-offs into the conversation before decisions are made. Microblog @antgrasso #EnterpriseAI Consider a common situation. Sales wants to promise faster delivery, while operations is already working close to capacity. Both teams may be making reasonable choices, but they are looking at different parts of the business. A decision coherence system can surface that dependency before the promise reaches the customer. It shows where assumptions differ and which constraints need attention. Outcomes from previous decisions can also guide what the organization does next. People still make the final choice. AI provides a common view, so local priorities do not create problems elsewhere. For this to work, organizations need consistent data definitions and clear responsibility for the information feeding the system. Otherwise, they may coordinate decisions more quickly without making them more coherent. Decision coherence does not require everyone to agree. It requires everyone to begin from the same business reality.

To be a freelancer, you need clients, plain and simple. Delano Saporu would say finding clients - and finding the right ones - is what you need to put first when you start out on your freelance journey. Read advice from Delano and 29 other freelancers: ow.ly/ekZK50ZRyRL

Export your AutoScientist open weights to Kaggle 💻 Can't wait to see what the community builds and shares!

Adding more agents is easy. Coordinating their work is the harder problem. The shift is letting specialists work in parallel, share context, and bring people in when judgment or approval is needed. That feels much closer to real teamwork. @hyperagentapp #HyperagentPartner

💡 What if you could own just a fraction of a building, a brand, or even art? That’s the promise of asset tokenization—turning ownership into something digital, flexible, and accessible. Follow @LindaGrass0 for more insights 💡

AI can change faster than an enterprise can redesign a process. That makes repeatability a strategic advantage: the less you rebuild from scratch, the more likely innovation is to reach production before the next wave arrives.

AI can change how work gets done faster than organizations can change around it. The challenge is closing the gap between what the technology makes possible and what the organization is actually set up to support: mck.co/4yJsyS6

How Does Physical #AI Change Embedded System Design? buff.ly/qMsW7JC @ElectronicDesgn #DigitalTransformation Cc @DeepLearn007 @GeoffAlexander1 @ahier @jblefevre60

You and your AI do not need to split the work 50/50; what matters is the right contribution at the right moment, not an equal share. Good collaboration is not measured by symmetry, but by whether each side contributes what the task needs when it needs it. Microblog @antgrasso In practice, the balance can shift from one task to the next. A procurement team may let AI scan thousands of supplier records and flag unusual changes. Most of the volume sits with the machine. Then one case reaches a buyer who knows that a supplier is going through a temporary production change. Context changes the interpretation, and one human judgment can outweigh hours of machine processing. AI may then return to the workflow and handle the next batch. No symmetry required. Responsibility follows a different rule. If an AI-supported recommendation affects a supplier decision, the organization still owns the outcome. So do not measure collaboration by how much work each side does. Ask whether AI is handling the work it can do well and whether people enter where judgment is needed. A 50/50 split looks neat on a slide. Real collaboration follows what the work requires. #HumanAICollaboration #FutureOfWork

Travel payments are more complex than they look. @bookingcom's approach shows how simplifying that complexity can make the customer experience feel seamless. mck.co/4h3dsRA