Four Perspectives on Research
Paul V. Mockapetris
Introduction
• My first second talk on the subject
• Institution & individual
• What’s important?
– Ideas & implementations that “change the world” in large or small ways
Why four perspectives?
• University & Student
• Research Institute & Researcher
• Public Institutions & Public Servants
• Startups & VCs/entrepreneurs
MIT years (Professional Student)
• Physics was it at the time
• Physics teaches the importance of
picking the right frame of reference
• While attending Physics classes, I got educated at:
– Architecture Machine (now Media Lab)
– IBM on virtual machines
• We did world class research both places out of necessity, but didn’t publish –
understand your field, publish if you can.
UCI years (Professional Student)
• Circa 1970s UCI/ICS believed:
– Software good
– Hardware, systems bad
- probably ahead of its time
• UCI has arrived
• Its never been easier to do world class research
– Tools in reach – Just do it
– Start now
– The importance of publishing
• Opportunity at the intersection of
disciplines
ISI years (Professional Researcher)
• The “importance” of being associated
with a company or institution
– Contribution to the bottom line
– Overhead
• Traps:
– the three year retitle – The safe theme
– Tenure vs hired help
ARPA/IETF (Public Service)
• ARPA
– Genetic diversity for funding with NSF, NASA, …
• IETF
– Research by committee
– The dilemma of the atomic scientists, ICANN
• Be a Program Manager?
– Best course on writing proposals – Best survey course
ever
– Set one or two agendas, deflect many by small amount
Venture Years
• News:
– Good: it really is a cycle – Bad: theme changes
every cycle
– The odd contribution of venture funding to open source
• Inventory overhang
– VC firms, niches
• Where is research done?
– Service companies ?
• VC challenges
– Ten year hangovers – Cash as leverage
• Entrepreneur
– VCs still need deals – Know the customer’s
needs
– Watch the clock
Where to look for Opportunity
• Less competition
– Junction of two disciplines
– Legacy researchers often do legacy subjects – seek new ground.
• The answer has to matter
Tactics
• Dealing with research sponsors
– Don’t expect to outsmart them easily, they do this every day.
– Building infrastructure is a necessary research evil, not usually a goal per se.
– Describe the experiments you will run and why they matter.
– Remember they need you if, and only if, you fill their needs.