Summary of "The Art of Doing Science and Engineering"

Core Idea

Vision drives linear progress; drifting produces √n progress—define your own goals of excellence and prepare continuously for opportunities
Master fundamentals deeply, not surface facts—depth enables you to recognize patterns others miss and adapt as knowledge doubles every ~17 years
Study method, not just results—how great scientists solved problems matters more than what they discovered

Schedule regular "big picture" time (e.g., Friday afternoons) away from daily tasks to reflect on larger directions
Hold 10-20 important unsolved problems in mind constantly; drop quickly if sterile
Ask "What's the real problem?" before solving symptoms—reframe to unlock better solutions
Anticipate technological shifts before they arrive; prepare your mind for changes you can't fully predict

Distinguish between education (what, when, why) and training (how)—both are essential
Learn the jargon of other fields but resist letting it limit your thinking; use as a tool, not a prison
Don't become an expert drag on progress—stay aware that what made you successful becomes counterproductive
Keep current constantly; experts who don't adapt are left behind

Simplify problems to fundamentals before adding complexity; remove extraneous details
Question accepted doctrine; verify that "impossible" claims still hold given changed circumstances
Great innovations come from outside the field—stay open to "crackpot" ideas from outsiders
Saturate your subconscious with important problems; temporary abandonment often precedes breakthrough insights

When mechanizing processes, design an equivalent product, not an identical replica
Build feedback into designs to tolerate inaccurate components; let feedback compensate for errors
Design systems for graceful degradation, not just hitting specifications—over-optimized designs fail when overloaded
Never optimize components in isolation; you will degrade overall system performance
Build flexibility into designs; anticipate changes you can't foresee

You get what you measure—choose metrics carefully; poorly chosen measures create perverse incentives
Question all claimed accuracies; independently verify data reliability (90% of measurements fall outside stated confidence limits)
Watch how measurement systems affect behavior; install safeguards against gaming the metrics
Question simulation outputs; large computers and pretty printouts don't guarantee validity

Sell your ideas three ways: formal presentations, written reports, informal conversations—all three mandatory
Give credit freely and share ideas openly—ideas spread your influence; secrecy restricts it
Help others succeed; recognition follows naturally from cooperation
Choose important problems worth solving; most scientists waste time on trivial work
The effort matters more than the result; the person you become through striving is the true gain

You are responsible for your decisions—don't defer accountability to committees or authority
Change yourself deliberately; you are the sum of your habits; small habit shifts compound over decades
Plan your future explicitly; drifting produces √n progress, vision produces linear progress
Don't become essential to your current role; that blocks promotion and locks you into the past

This week: Define 3 clear personal goals of excellence; schedule recurring "vision time" (minimum 2 hours/week) for strategic thinking
This month: Identify 10 important unsolved problems in your field; study how 3 great practitioners solved similar problems (method, not just results)
This quarter: Audit your measurement systems—are your metrics creating perverse incentives? Redesign one metric to reward the behavior you actually want
Ongoing: Master one fundamental principle deeper than peers; learn jargon in adjacent fields; question one "impossibility" claim with fresh assumptions