In 1971, the economist and cognitive scientist Herbert A. Simon wrote a line that explains the world we now work in better than most things written this year:
“A wealth of information creates a poverty of attention.”
— Herbert A. Simon
He wrote that 54 years ago, long before the internet, search engines, or AI assistants. His point was simple. Information consumes something. What it consumes is the attention of the people who receive it. So when information becomes abundant, attention becomes scarce. And in economics, the scarce thing is the valuable thing.
Today AI hands every one of us almost infinite information for free. Ask a question and you get a complete, well-formatted answer in seconds. The supply of information has effectively become unlimited. By Simon’s logic, that means the value has moved. It is no longer in the information itself. It is in the attention required to notice what matters inside it.
Why this matters for engineers in the AI age
For a long time, knowing things was a real advantage. The engineer who had read the datasheet, memorised the API, or remembered how the scheduler behaved was genuinely more useful than the one who had not. Recall was a moat.
That moat is shrinking. A language model can recall almost any documented fact instantly and without error fatigue. If your main professional value is that you can remember and retrieve information, you are now competing directly with a tool that does it faster and cheaper.
But there is a part of the work the model cannot do for you. It cannot decide which of its ten suggestions is correct for your hardware. It cannot notice that the answer it gave is confidently wrong because it did not know your constraint. It cannot tell that the real problem is not the one you asked about. All of that requires attention and judgment applied to the specific situation in front of you. That is the part that stays valuable, because it is the part that does not come in the download.
So the skill that matters is no longer “how much do I know”. It is “how well can I attend to what is actually here, and judge it”. In an information-rich world, that is the scarce resource Simon was describing.
How to build attention as a deliberate skill
The encouraging part is that attention is not a fixed trait. It is a habit you can train, and most people around you are not training it. A few concrete practices that work:
- Read the problem before you read the answer. When you hit an error or a task, read it slowly, twice, before you reach for AI. Make sure you understand what is actually being asked. The answer is cheap; understanding the question is the work.
- Single-task on hard work. When you are debugging or designing, close the extra tabs and put the phone away for a fixed block of time. Divided attention feels productive but it is the state in which you read a line of code without seeing it.
- Generate before you look up. When you are learning, attempt your own answer first, even a rough one, then let AI check and correct it. You keep the speed of the tool without giving away the effort that actually builds the skill.
- Verify AI output on purpose. Treat every generated answer as a draft to be checked against your own knowledge and the real system, not as a fact to be pasted. The verification is where your judgment grows.
- Spend your attention on what is scarce. Let the tool handle recall and boilerplate. Spend your own attention on the parts that need a human in your exact situation: the trade-off, the edge case, the assumption nobody stated.
None of this means avoiding AI. The goal is the opposite. You want to use the tools fully while keeping the underlying skill that makes you able to tell when the tool is right and when it is not. Outsource the retrieval. Keep the attention and the judgment.
The simple takeaway
Simon’s sentence is a quiet instruction for a career in the AI age. When information is everywhere and free, do not compete on having more of it. Compete on attending to it better. Notice what others skim past. Read the thing that matters with your full mind. That trained attention is the moat that AI does not remove, and it is the difference between using these tools and being replaced by them.
At TECH VEDA, this is why our training is built around working through real problems with full attention, not just collecting answers.
Source: Herbert A. Simon, “Designing Organizations for an Information-Rich World” (1971).
— Raghu Bharadwaj
