Intellectual Frontiers Network

How a search runs, thesis to outcome

A search starts with a written thesis, so when it ends we can see where the thesis was wrong. In between, a person makes every judgment call.

Most search starts with a query and ends with a list. That puts the judgment on you. We do the judging first: a person reads every packet, decides who is worth contacting, and answers for the decision in writing. When the search ends, the outcome gets compared to the thesis, and the next search starts from the difference.

The stages

  1. Stage 1

    Thesis

    Someone writes down who would have to exist for a piece of work to succeed, and what would make that person unusual.

  2. Stage 2

    Discovery

    We look outward at what is already public. The Internet is the database. We do not copy it into ours.

  3. Stage 3

    Surfaced

    A small set of candidates comes back, each one attached to the thesis it answers.

  4. Stage 4

    Evidence

    Every candidate gets a packet: what was found, where it came from, and what we still do not know.

  5. Stage 5

    Human judgment

    A person reads the packet and decides whether to reach out. No score decides this.

  6. Stage 6

    Contact

    One specific message about one specific thing, sent by a person who can act on the answer.

  7. Stage 7

    Conversation

    What we learn in the conversation is written down, because it is the part the Internet does not have.

  8. Stage 8

    Validation

    The hypothesis about what this person is unusually good at either holds up or it does not.

  9. Stage 9

    Match

    The person is connected to the hunt, the venture, the fund, or the editorial project that needed them.

  10. Stage 10

    Experiment

    The match starts as a small, bounded piece of real work rather than a title.

  11. Stage 11

    Outcome

    The experiment produced something or it did not, and we record which.

  12. Stage 12

    Learning

    The thesis is corrected with what actually happened. The next search starts from what this one got wrong.

A failed search is still useful to us, because it tells us what the thesis got wrong. That is why the second search for a similar person is almost always better than the first.

See the hunts running now