Back to more

Notes

The Art of Doing Science and Engineering

Notes from The Art of Doing Science and Engineering by Richard Hamming.

Greatness

  • An image of greatness and its unapologetic pursuit is allowed. It is cowardly not to.
  • Lunch with anyone he could learn from or argue with.
  • Greatness is a practice, not a trait. It is not only to be admired, but to be aspired to.
  • "Luck favors the prepared mind."

Orientation

  • The study of successes is as important as the study of failures.
  • There are many ways of being wrong and one of being right.
  • Mull things over. Compare what I say with your own experiences.
  • Style: study under a master, then forge your own from the past plus your own abilities.
  • Meta education is what matters.
  • Expect the constant need to learn new things.
  • Learn the fundamentals. They should let you derive everything else in the field.
  • Science: if you know what you are doing, you should not be doing it.
  • Engineering: if you do not know what you are doing, you should not be doing it.
  • Short-term predictions are usually optimistic; long-term predictions are usually pessimistic. People cannot comprehend the power of compounding.
  • In AI, almost every leader has made predictions that did not become true, and may not become true in your lifetime.
  • "You must try to foresee the future you will face." A career with a vision moves with distance proportional to n; no vision moves with sqrt(n). Accuracy is not the main point.
  • He spent 10% of his time, Friday afternoons, thinking about what would happen.
  • Ask what is possible, what is likely to happen, and what is desirable to have happened.
  • Having a vision is what separates leaders from followers.
  • Computers are cheaper than humans, faster, more accurate, more precise, more reliable, easier to retrain, and can work in hostile environments without boredom, personal problems, or egos.
  • You should try to make significant contributions to humanity. The aim itself is worthy. A life without struggle is hardly a life worth living.

On Foundations

Computers make it possible for robots to make many things, not only through standard von Neumann computers, but through neural nets and fuzzy logic.

Creativity

  • Combining three extensively developed fields simply would be a large creative act.
  • Creativity does not depend on how hard the act was.
  • "I believe the future will have a much greater need for style."
  • Creativity is like sex: you can read all the books, but without direct experience, you know little.
  • Pattern:
    • Recognition of a problem.
    • A long period of refinement, emotional involvement, and intense thinking. This can produce a solution or abandonment. Temporary abandonment is common; single-minded pursuit often does not work because the subconscious needs room to find a solution.
    • A moment of insight. You see it. Usually you are wrong, but maybe the problem has to be altered to fit the solution.
    • Ask: "If I had a solution, what would it look like?" What must it involve? This is the reversion thinking pattern.
  • Manage the subconscious: saturate it. Think seriously about nothing else for hours or days. Deprive it of anything else.
  • "Luck favors the prepared mind." Think constantly and occasionally you are lucky.
  • You need hooks, analogies, and reminders: x reminds you of y.
  • Build an information retrieval system. Make a habit of turning over new information and creating hooks for retrieval.
  • Mold new ideas by trying to connect them to relevant information.
  • When you learn something, think of other applications for it, maybe not from the past, but ones that might be found in the future.
  • He found luck by making absurd timelines and almost being late. Pride and self-confidence mattered.
  • Dropping a problem: previous successes convince people that they can solve any problem.
  • Great things might arrive early. There is no perfect moment, and it will never arrive.

Experts

Most great innovations come from outside the field, not from insiders. If an expert says yes, it is yes. If he says no, get a second opinion.

Data

Never process any data until you carefully examine it for errors.

Systems Engineering

People get too deep into the details and forget the bigger picture. The goal is not to give a lecture, but to educate. Avoid myopic views. Larger pictures matter, though some larger pictures are wrong.

You Get What You Measure

  • Recruiting for research means looking for originality in science and engineering. Originality can also show up in behavior and dress.
  • At Bell Labs, research people went out to do the hiring for research.
  • Even in entertainment there are first and last performers. Society rates people, and in a fast-changing society this happens often.
  • Ask whether things are as reported. Ask questions.

You and Your Research

"Yes, I would like to do great work." Pick your own goals. What great means is up to you, but have them, and make them high.

Luck and Preparation

  • Objection: great science is not done by luck.
  • But luck favors the prepared mind.
  • Newton: "If others would think as hard as I did, then they would get similar results."

Courage and Confidence

  • Great work is something else than mere brains.
  • One success can bring confidence and courage.
  • Once you get your courage up and believe you can do important problems, then you can.
  • "You would be surprised, Hamming, how much you would know if you worked as hard as he did that many years."
  • The more you know, the more you learn; the more you learn, the more you can do; the more you can do, the more opportunity opens up.
  • If you believe too much, you will not notice flaws. If you doubt too much, you will not start. Ambiguity is important.
  • If you are completely submerged in a problem, your subconscious will find answers.

Important Problems

  • "What are the most important problems in your field?"
  • You will not do important work otherwise.
  • Great scientists have a set of 10 to 20 important problems in their field, always looking for an attack.
  • Important means there has to be an attack as well.
  • Have "great thoughts time" only to discuss these things.
  • When you see a new idea, ask whether it attacks one of your important problems. If yes, go for it.
  • Working with the door closed gets more done, but leaving the door open occasionally gives clues about the world and what might be important.

Building and Selling

  • Some people do slightly the wrong thing, not much, but enough to miss fame.
  • Do your job so others can build on top of it.
  • You have to sell the work, not only do it.
  • Next time you open a journal, ask why you read some papers and not others.
  • Write clearly, give reasonably formal talks, and learn to give a conference talk.

Working With the System

  • Good scientists often fight the system instead of learning to work with it and use what it offers. This does not work.
  • Make an effort to appear to conform better. Dress according to audience expectations.
  • Or fight it steadily as a small undeclared war for your whole life.
  • "Which do you want to be? The person who changes the system or the person who does first-class science?"

Emotional Involvement

  • If you do not get emotionally involved, it does not create stress. You need stress.
  • Pick capable people to have conversations with.