Unworkable Ideas

October 29, 2025 | hack writer

Since When Is Artificial Good

Exploring why human judgment remains indispensable in an age of AI.

Real ingredients are good. Artificial ingredients are bad.

Except when it comes to intelligence. We recoil at synthetic flavors, preservatives, and additives — yet we celebrate artificial intelligence, trusting machine-generated decisions without hesitation. Why do we fear chemistry in our food but embrace it in our thinking? This paradox is worth examining.

We live in a world governed by rules enforced programmatically. Most of the time, these systems work beautifully. But edge cases emerge — messy, human situations that require judgment. That’s where AI stops, and humans must decide.

Artificial Intelligence is impressive, but it is no substitute for real intelligence. AI assumes the future will resemble the past. It predicts outcomes based on historical trends. But what happens when patterns break — a market shifts suddenly, a legal precedent changes, or a competitor introduces a disruptive innovation? AI will suggest plausible options, but the correct judgment often lies outside its model.

Real Situations Require Real Consideration

Think of a baseball manager deciding whether to pull a pitcher late in the game. Data provides probabilities, but other factors — fatigue, weather, the pitcher’s body language — matter. The manager weighs options and decides. Sometimes they’re right, sometimes wrong. That judgment is the essence of intelligence.

Or consider a manager evaluating a high-risk client contract. Data may suggest it’s safe, but hidden contextual factors — prior client behavior, market volatility, or internal team readiness — require nuanced judgment. The right choice often depends on experience and intuition, not historical patterns.

Robots can call balls and strikes, or flag contracts for review. But they cannot weigh the full complexity of the moment or anticipate consequences humans understand instinctively.

Rules, Judgment, and Chance

Rules create order.
Chance introduces uncertainty.
Judgment makes meaning in between.

Even the best rules cannot cover every situation. Random events happen. Interpretation is required. Human judgment remains the differentiator in complex systems.

Of course, not everyone uncritically worships AI. Skeptics question bias, accountability, and over-reliance on algorithmic advice. Yet the general narrative often leans toward excitement and adoption, sometimes without reflecting on what human judgment might be lost in the process.

Key Takeaways

AI amplifies pattern recognition but cannot replace human judgment in novel or messy situations.
Real intelligence relies on context, interpretation, and intuition, not just data.
Competitive advantage — in business, law, sports, or life — comes from understanding nuances that machines cannot see.
Human judgment should be protected, prioritized, and cultivated deliberately; otherwise, intelligence risks becoming artificial in all the wrong ways.

October 25, 2025 | hack writer

Losing the IT Productivity Battle

Information technology is not really about productivity anymore, it is really about self perpetuation and expansion. It is not about users and customers anymore, if it truly ever was.

Sure, there are PCs, Macs, Linux, Android, iOS, and a dozen other systems. They’re smarter, faster, more connected than ever — and none of it matters. We’re reaching a point where the value created by all this power will be overshadowed by the cost of maintaining it.
We’ve got databases, knowledge bases, big data, training data, synthetic data — and honestly, none of it really matters.

Look at a typical Managed Service Provider: ticketing, asset management, billing, CRM, quoting, documentation, reporting, planning. Most of these systems come from the same vendor. They’re supposedly integrated. And yet—contradictory data, incomplete records, gaps everywhere. The promise of seamless integration delivered as distinct data sources that talk at each other, but don’t really understand each other.

I cringe at the sheer volume of information bloat. I hate that I’m complicit in it — building tools to manage tools to manage data no one even reads.

We make tools to manage tools.
Reports about tools.
Reports about the reports.
It’s turtles all the way down — only digital.

There are whole industries dedicated to making tools to make reports about tools. Each one promises clarity, but delivers opacity. Every addition has a plausible justification, and every new layer adds just enough complexity to require another layer.

You get lost in the weeds of tools and data until you can’t see the absurdity anymore.

The digital minimalists are onto something, but the war was never just about attention — it’s about the collapse of meaning itself. Information abundance has created a new kind of poverty: a poverty of sense, of coherence, of purpose.

Someone always has an advantage when information is asymmetric. That’s not new. But now the asymmetry isn’t just between people — it’s between systems and everyone else.

We’re not thinking clearly because we’re still trapped in the old paradigm, patching and tweaking what exists, pretending it can evolve into something new. But this new paradigm is an illusion — temporary at best.

It’s easy to see AI’s current inadequacy and assume human judgment will always matter. It’s just as easy to believe AI will inevitably surpass it. Both views miss the point: we’ve already outsourced our judgment — to the systems we built to manage ourselves.

And now, a new phase is beginning — the rise of the agents.

Software that acts on our behalf, deciding, negotiating, filtering, buying, and prioritizing — all without us touching a screen.

Maybe that’s the real lost war of information technology — not the fight against chaos, but the quiet surrender of agency.

It sounds efficient, even liberating. But it also means we’ll have even less visibility into the flows of information that shape our world.

October 11, 2025 | hack writer

The Great SaaS-Lighting: How IT Users Got Gaslit

Pay as you go
Just pay for what you need
Free up time
Free up capital
Focus on your business not the technology

But that’s not quite the way it worked out, is it?

Maybe they meant it when they said it, but it is not the driving force behind the great SaaS purveyors of our day. Yeah, I am looking at you Microsoft, you too Google, and where do you think you’re hiding Intuit? There is nothing wrong with creating a product that your customer wants to buy, but something is off when the customer is forced to buy something they don’t want. Customer needs are completely secondary to customer lock-in. Sadly, the SaaS model has become too big to care about customers. They spew customer satisfaction surveys at you after every interaction in an effort to show that they care. But these surveys are just another brick in the big data wall. The results are secondary. They kind of care what you think…but mostly they just need you to hang around and keep paying. They collect the data to guide incremental improvements around the edges.

The irony is that just about every SaaS vendor has created the role of customer success manager. These are people that are assigned to your account to help you onboard and have success with the product to keep you from off-boarding. Success doesn’t necessarily translate into helping your organization succeed, just that you ‘succeed’ enough with the product.

I don’t begrudge anyone’s success in creating a product that customers find useful and want to buy. That is not an easy thing to do. But after a while, the SaaS business model is not about customer success or satisfaction anymore. It is about customer submission and inertia. At a certain point, the customer base becomes too big and the product becomes too big to change.

Safety in Numbers

It is the path of least resistance. If everyone else is doing it, it must be good or at least good enough. Plus, there is value in the network effect. Yet, numbers are only safe up to a point. Numbers blind you to unseen risks. Black swan type risks. Rare, but catastrophic. They are so rare and potentially disastrous that no one really thinks about them.

Maybe your backup, disaster recovery, business continuity or whatever they are calling it these days will save you from a single system failing, but it won’t protect you from context and know-how loss. The real problem isn’t loss, it’s accumulation. Too many programs, too many APIs, too many integrations, too much complexity masquerading as sophisticated systems. Context recovery systems don’t exist, yet successful organizations rely on context, not data. Terabytes of data mean nothing without knowing why you have the data, what it means and how you need to make use of it. Maybe the software rules take care of it, but that is a dangerous thing to rely on.

Information and content is infinite and stochastic. It is not necessarily predictable. More information doesn’t lead to better decisions, it just leads to more data.

This preys on the fear and risk of not knowing.

You can never know all the information before you have to make a decision, but when faced with unknowns and uncertainty, adopting “best practices” provides a cozy security blanket.

Undifferentiated Best Practices

Gotta love the industry ‘best practice’ templates. About a million years ago, I remember an old printed newsletter called the ‘Best Practices Report’ which featured, you guessed it, best practices. The trouble with best practices — then and now — is that they pretend that the world has stopped changing. But the reality is, the world is not static. Things change. You need to keep getting better, you need to keep evolving. Blindly adopting templated best practices is not a path to be best (to paraphrase someone), but rather a path to bland mediocrity.

This preys on the ‘why reinvent the wheel’ logic. Why spend the time and effort on figuring out something that has already been solved.

The reality is that you will be really good at achieving parity with your competition.

Bland and Generic Applications

Speaking of bland mediocrity, all you have to do is to look at the landscape of commercial software. There are thousands of applications across thousands of categories, but we are still getting different takes on note taking or calendar applications.

Some programs might look prettier or feel more intuitive, but they are tackling the same problems.

Software keeps iterating on solving the same soluble problems because the remaining challenges are really difficult to solve with technology. Communication and coordination are full of nuance and subtleties that defy digitization.

This is not unique to SaaS vendors, but most, if not all, software vendors have jumped on the SaaS bandwagon for marketing and selling their products. The free version that does just enough to lure you into the paying membership. Then you face the standard three options to subscribe with the good, better, best offers.

Adding communication tools hasn’t improved the quality of communication despite massively increasing the volume of communication.

Let’s Go to the Mall

SaaS has become the technology American shopping mall of the 1980s. It is overpriced and predictable. The goods are largely the same in every mall. This is not a dynamic market. It is very much a controlled market. The landlord sets up the platform and the retailers rush in to this great location to make the huge profits and get the advantages of scale. The retailers that can afford the mall’s rates, have very controlled experiences. The mall of the 1980s was not a place of bold experimentation and risk taking. The risk was signing the lease.

While Google and Microsoft are stores in the mall, they are also the landlords. They control the mall experience. Apple runs its own mall — just shinier, not different. (Today’s physical malls would be ghost towns without an Apple Store bringing in traffic.)

Somewhere along the line, the culture tired of the mall experience. Across the country, the US is full of abandoned malls. The model works up to a point and then the fashion changes.

The small store with the carefully curated merchandise appears on the scene and draws in its crowd.

The future is much the same for Information Technology. The point is not to have the same system that everyone else does. The point is to have the information system that works for you.

This issue of the newsletter was written on a self hosted WordPress site on multiple devices with no monthly fees.

June 28, 2025 | hack writer

The Ridiculous Blog

Hundreds of posts on a topic that apparently no one gives a fuck about. It is only the most important topic in the history of the world. The discovery, transmittal and use or misuse of information. What could be more important than that? At least that is what he told himself over and over again. What he failed to consider is that people take information for granted and they just don’t have any time. Who really gives a fuck and who has the fucking time? We are overloaded with information anyway. We are always taking shortcuts. Give me the most amount of information that I can use and do it in the shortest amount of time.

The attitude is don’t waste what is left of my life with your tedious, ponderous and ultimately pointless alleged insights. Nobody needs it. Just give them information in short bits and keep them amused.

Yet for some fucking reason, he kept writing post after post. Following blog posting best practice after best practice. After a while, he was writing his blog to please the alleged rules of search engine optimization that all of the so-called gurus and easy to use apps tell you to follow.

But who was this blog for? IT people? Surely not. It is not what they do. Management people? Maybe but they tend to defer to technology people on technology issues. The blog was for himself mainly, but it was anything but therapeutic. It just made him more and more alienated from everything.

It was so clear to him, but everyone else was lost in their own delusions. Yes, I can integrate 365 with Salesforce and I will be able to track every interaction with our customers and prospects. Every email. Every phone call. Every note to self. It will be so glorious, yet so pointless. What can you possibly do with all that information?

He knew the harsh reality. The value of the information is more limited than you think. Maybe you will sell more, maybe the customer relationship will be that much richer. But is that really enough? Is your understanding of your customers that much better. Does it justify the effort involved in collecting, storing, securing and archiving all that information.

Even with all the information in that database, you still get things wrong. The information itself represents just a moment in time. You will still have misperceptions and miscommunications. That is the one thing that is guaranteed in a world of infinite information and infinite possibility. The infinite opportunity to screw up royally.

On the surface the cost of extra information is near zero, but the cost of dealing with extra information is hard to calculate.

June 27, 2025 | hack writer

Noise Plus Interruption Doesn’t Equal Information Management

The formula for effective information management is a little more nuanced than that. Survival in the real world often requires the ability to block out distractions and focus on what really matters at the moment.

When you are being chased by a lion, the only information that matters at that time is information that helps you escape. This is not the time to stop and look at a billboard ad for sneakers that help you run like the wind. While you may wish you had those sneakers, that distraction is not useful at this moment. The key at this point is focus and not data accumulation.

However, in non-crisis situations, data accumulation is usually considered a good thing. More information makes us smarter. While this is true in theory, information is only helpful if you are able to use it in a way that gives you some advantage at some point. However, the problem with information is just the sheer volume of it especially as it evolved over time. Information is not static. What was true in one circumstance at one moment in time may not be true in another similar instance in the future.

It is great to build up a knowledge base that you can draw on, but the human brain is the ultimate information filter. It can only process a finite amount of information at a time. The same is true of even the most powerful computers. They are only processing a fraction of the available data at the time working with all kinds of human simplifications and assumptions built in. It is no wonder that so many simulations and forecasts turn out wrong.

The answer is not more information and more distractions, the counter intuitive solution is to allow more time to focus.

June 27, 2025 | hack writer

Optimization Ruins Everything

Optimize this process, optimize that. Optimize for X. Sometimes it is just a waste of brain power and energy. It doesn’t really matter what you are trying to optimize, the process of optimization reduces the potential of variability. While variability can produce seriously negative consequences, it can also unleash creativity and delight.

Reducing waste and eliminating extraneous steps are noble goals, but digitizing and homogenizing everything ultimately leads to something that lacks meaning.

It becomes the digital equivalent of Tiny Boxes. A digital landscape of generic forms designed to streamline some backend process that provides little benefit to the end user.

The forms are almost always clunky and they are either too complex or too simplistic to convey the valuable contextual information. I have been on both sides of the form and have been guilty of wanting to use them to streamline and simplify things.

And don’t get me started on automated call centers and the call routing systems. In theory this is intended to optimize call handling and routing, but it may actually be optimized to frustrate and annoy any inbound callers. The menu options are never quite what you are looking for. The voice recognition doesn’t work reliably, you need to repeat the reason why you are calling at each handoff and you dread the thought of ever having to call in again. Coincidentally, it seems that everyone is simultaneously changing their menu options so you are always encouraged to listen carefully.

Finally, there is the support ticketing system which is a beautiful hybrid of the two ‘optimized’ processes above. Tickets and issues that are force fit into specific queues which may or not be properly categorized or may span across multiple categories. The ticket notes get longer and longer and more confusing as more people are brought in to handle the ticket. The rationale of a ticketing system makes sense, but it is not necessarily optimized to help the person submitting the ticket or the help desk for that matter.

Maybe a little less effort on optimization of the system with a little more focus on optimizing the user experience would produce more lasting results.

Optimization in and of itself, may be a dead end.

June 25, 2025 | hack writer

The Most Unworkable Idea of All: Centralized Control

On the surface, centralized control makes all the sense in the universe. Unfortunately, the universe has an infinite amount of sense and an infinite amount of nonsense.

A world with centralized control is theoretically better prepared and more efficient, but managing a world with infinite variables is impossible.

Still we try to control as much as we think we can. There are some domains that we recognize as out of control, but with information we really do try to control it. We want systems to capture, manage, process and retrieve information. Truth be told, we want the right information to guide our decisions and in many cases, we don’t necessarily want others, especially our competitors to have access to this information that we feel is advantageous.

We have all kinds of unconscious information behaviors. What we take in, what we remember, what we share. Humans are constantly processing information and probably some information behaviors are through evolutionary biology. Teaching our children, learning to collaborate, starting to keep records all seem to encourage survival and reproduction.

There are clearly good reasons to manage information, but it is impossible to manage it all. While having more information provides more confidence, it doesn’t necessarily lead to better outcomes. You need to have the right information at the right time to have any kind of advantage, and you can’t predict what information that you will need.

Humans may anticipate needing some of the that they will need, but you can’t know everything in advance.

Information systems try to provide this illusion of control of information, but there is still way more information and factors that are outside the boundaries of every information system ever developed.

November 7, 2023 | hack writer

Naturally Reductive

There are so many ways to be reductive. Reduction is so seductive. It is easy. But rarely does anything work quite that way. There are more factors and interactions at play than you can imagine.

But to function, the brain simplifies.

There is System 1 and System 2 processes that Daniel Kahneman

Humans need both.

The perfectly rational entity does not exist.

Even perfect rationality can be foiled by bad luck.

So where does that leave us.

Guessing. Ruled by emotion. Rationalized after the fact by selective information. And most importantly, pretty lazy.

November 7, 2023 | hack writer

Framing

Framing is the ultimate skill and it happens all the time. Things get framed for us and we also frame things on our own.

There is so much information to take in and process all the time that we use frames to help us make sense of things. Just like in photography, the framing can be somewhat arbitrary. What is included inside the frame versus what is left out makes a big difference.

Things happen inside the frame and things happen outside the frame.

Still, there are things inside the frame that we miss and there are things outside the frame that we convince ourselves are there.

As part of the human bag of tricks, our brains allow us to see things that aren’t there. This is helpful, but it also leads us astray. We often fill in missing pieces with things that aren’t there.

Timing is an overlooked aspect of framing. Each second in time consists of multiple frames. But even with an infinite amount of frames, it is impossible to catch everything.

Something is bound to go unnoticed.