NotePD Loader
Ideas Post

10 reasons why models and frameworks can be flawed

A framework is a general, reusable solution to a common problem within a given context. A model is a specific implementation of a framework. Models and frameworks are ways in which humans make sense of reality.

10 reasons why models and frameworks can be flawed

    1. Skin in the game

    Be skeptical of any model or framework that has not been applied or tested by the person espousing it. Or one that is developed by someone who has no stake, direct experience or knowledge of its end-users and customers.

    2. Smart people support bad models or frameworks

    We need to consider the vested interest of intellectuals or industry experts that are pushing a certain framework. Often it has more to with money, cognitive dissonance (the need to validate their own beliefs), securing their careers or maintaining a reputation. Rather than the efficacy or relevance of the framework.

    3. It's hard to predict the future even with the best model

    Look no further than the financial crisis of 2008. The quants or financial math experts built elaborate models for predicting the stock market. The models were built on flawed assumptions, namely that past behaviour always indicates future actions.

    The overleveraging of housing, underwriting loans to non-qualified people, rating agencies and a slew of other variables were not factored into these models.

    4. A framework is a helpful guide, but ultimately it comes down to trial and error & experimentation

    We like to follow frameworks because it gives us a sense of security. But coming up with a unique framework and testing it on unproven grounds is scary and risky. Often time we need to reevaluate our own framework repeatedly until we find one that works based on experience.

    5. The framework creator is not accountable for adverse impacts

    The formulators of frameworks are not accountable for it hurting society through unintended consequences. This is particularly true of think tanks and policymakers that define our laws and regulations.

    On a positive note, many AI researchers are cognizant of inherent biases found in AI models and at least address this problem.

    6. Use data driven models because they incorporate experimentation

    There is a strong push to make data-driven decisions in organizations and move away from strictly experience or intuitional frameworks. Because data can provide an objective view of what is going on in the world. For example, google search results are a strong indication of sentiment vs an informal opinion poll on a given topic.

    You may argue that AI models are also subject to bias and are by no means perfect. Fair enough, however, they do uncover patterns in our data which were not previously known and make us reevaluate tried and true frameworks.

    7. Frameworks are remixes or a patchwork of previous frameworks that were mildly sucessful

    Since the 90s a number of rehashed or slightly varying frameworks were created around process improvement have popped up: business process reengineering, six sigma, lean six sigma, and kaizen.

    Digital Transformation is just Business Process Automation + AB-Testing + Social Media Marketing + Customer Experience+Web3+Data Science mashed into one marketable umbrella term as a revolutionary way of doing business.

    In companies, you will have an old operating model recycled into a new operating model with small variations and rebranded as innovative.

    8. The collective knowledge of people outweighs a static framework

    The collective fragmented, mundane, experiential knowledge of a populace may be more valuable than a framework or model cooked up in a lab.

    Frameworks are often created in isolation as thought experiments that sound compelling. But carry a fatal flaw during implementation that I have seen over and over in corporations and new businesses: They don't engage with their end-users. They create solutions based on a framework but don't bother to talk to the people that use their solutions.

    9. Frameworks that are complex, interesting, original, persuasive, elegant or ingenious may not necessarily practical

    10. We tend to accept models and frameworks that confirm are preconceived belief rather then challenge them

0 Like.0 Comment
Eyegorand 3 more liked this
Comments (0)

No comments.