If you want to build real-world decision intelligence…

Posted on August 5, 2025 by decisionmanagementcommunity

Adam DeJans Jr. just posted on LinkedIn practical advice on how to build real-world decision intelligence: Most “decision intelligence” projects fail because they skip one thing: Engineering. A model is not a system. A dashboard is not a decision. If you want to build real-world decision intelligence, here are 5 practical principles I use when building production systems:

  1. Build around the decision, not the data
  2. Encode policies, not one-off answers
  3. Build the simulation before the pipeline
  4. Log everything
  5. Test for uncertainty, not just correctness

It’s not about perfect models. It’s about resilient systems that make smart decisions, log them, and learn. Test for uncertainty, not just correctness. Link

Posted in Decision Intelligence | Leave a comment

Challenge Aug-2025 “Grid Covering”

Posted on August 3, 2025 by decisionmanagementcommunity

This month, we offered our readers to ponder a problem that was among the most complex on the 2025 International Mathematical Olympiad. This problem did stump DeepMind and OpenAI’s models, but it wasn’t just problematic for AI. Of the 630 student contestants, 569 also received zero points. Only six received the full credit of seven points. Will our readers be able to solve it using their favorite tools or just pen and paper? Link

Posted in Challenges | Leave a comment

Measuring AI Impact on Experienced Developers

Posted on August 2, 2025 by decisionmanagementcommunity

“Surprisingly, we find that when developers use AI tools, they take 19% longer than without—AI makes them slower.” Link

Posted in Artificial Intelligence | Leave a comment

Interesting Kaddle Competition

Posted on July 28, 2025 by decisionmanagementcommunity

Markets don’t always stick to the plan. Like a sudden shift in the weather, price swings can throw off even the best predictions. Investors and businesses need models that can keep up, but many struggle to adapt when past trends no longer apply.

In this Kaddle competition, you’ll develop an ML model to predict commodity prices using historical data from LME, JPX, US Stock, and Forex markets. The challenge is creating a model that stays accurate even as conditions change. Total Prizes: $100,000. Entry Deadline: Sep 29, 2025. Learn more

Posted in Challenges | Leave a comment

Meta-Decision Support

Posted on July 26, 2025 by decisionmanagementcommunity

Adam DeJans Jr.: “What if we treated the design of metrics, thresholds, and goals as decisions that need feedback loops and learning just like any other policy? That’s where I think the future lies. Meta-decision support: helping leaders choose what problems to define, what metrics to use, and how to revise those as the environment changes.” Link

Posted in Decision Modeling, Decision Monitoring, Decision Optimization, Supply Chain, Uncertainty | Leave a comment

Writing is thinking

Posted on July 24, 2025 by decisionmanagementcommunity

This article discusses the value of human-generated scientific writing in the age of LLMs. “Writing compels us to think — not in the chaotic, non-linear way our minds typically wander, but in a structured, intentional manner. By writing it down, we can sort years of research, data and analysis into an actual story, thereby identifying our main message and the influence of our work.” Link

Posted in Thinking, Writing | Leave a comment

A call for AI substance not for more hype

Posted on July 24, 2025 by decisionmanagementcommunity

Today Stéphane Dalbera posted such a call: “… in the case of LLMs, too often the discourse remains high-level and aspirational. Demos abound, wow effect, but operational insights remain scarce. For a field that claims to redefine productivity, creativity, and even human reasoning itself, the lack of grounded, transparent, and reproducible narratives is troubling.” Interesting quotes provided in comments by Vincent Lextrait: Link

Continue reading →
Posted in Artificial Intelligence | Leave a comment

“Decision Intelligence is AI for Grownups”

Posted on July 19, 2025 by jacobfeldman

This phrase is often used to highlight how Decision Intelligence (DI) represents a more mature, pragmatic, and business-focused application of artificial intelligence. I asked Copilot why some people describe DI that way. See the answers at https://www.linkedin.com/feed/update/urn:li:activity:7352416233801060352/

Posted in Artificial Intelligence, Decision Intelligence | Leave a comment

Is what photography did to painting similar to what AI does to software development? 

Posted on July 14, 2025 by decisionmanagementcommunity

Google this question and see different answers. “This industry, by invading the territory of art, has become art’s most mortal enemy.” Was this quote said in 1859 about photography by Charles Baudelaire or in 2025 about AI art by somebody else? Read “Photography was historically considered art’s most mortal enemy. Is AI?” by Lizzy Larson.

One of the answers was published yesterday by a fine art photographer, Craig Boehman, and is called “In Defense of AI Art: History Repeats Itself, Again, Again, and Again“: The “anyone can do it” and “it will put people out of work” arguments. And has painting been superseded by photography? What nonsense. Check the auction houses. Paintings go for hundreds of millions of dollars and a mere photo has yet to reach 10 million. And there are very few cases of photos being valued this high. Naturally, value isn’t the only way to judge an art form. But there is no evidence whatsoever that photography has overtaken painting.

Posted in Art, Artificial Intelligence, Software Development | Leave a comment

The Rise of Small Language Models (SLMs)

Posted on July 11, 2025 by decisionmanagementcommunity

Specialized small language models (SLMs) can outperform large, generalist models. Want to learn how? Read Armand Ruiz’s post on LinkedIn: “Inference is cheaper. Iteration is faster. Fine-tuning takes hours, not weeks. SLMs can run locally, privately, and securely and no datacenter needed“. See the related research paper stating that “SLMs are sufficiently powerful, inherently more suitable, and necessarily more economical for many invocations in agentic systems, and are therefore the future of agentic AI“.

Posted in Gen AI, LLM, SLM | Leave a comment