Data Relevance Quotes

One thing ImageNet changed in the field of AI is suddenly people realized the thankless work of making a dataset was at the core of AI research.

Most of 'big data' is a fraud because it is really 'dumb data.'

Information can tell us everything. It has all the answers. But they are answers to questions we have not asked, and which doubtless don't even arise.

Our criteria for good data feeds are that they need to be dynamic enough to be interesting, something that users want to attune to throughout the day, etc.

Data isn't information, any more than fifty tons of cement is a skyscraper.

Measurement is fabulous. Unless you're busy measuring what's easy to measure as opposed to what's important

Data is like garbage. You'd better know what you are going to do with it before you collect it.

Do not seek for information of which you cannot make use.

Information is a difference that makes a difference.

Good data is hard to find. Hard data is good to find.

All information is good, even when it is bad.

The more up-to-date the information, the more valuable it is.

Any information is valuable to the degree that you can use it.

You can have data without information, but you cannot have information without data.

If the metrics you are looking at aren't useful in optimizing your strategy - stop looking at them.

Results "are no good unless they answer (or can be made to seem to answer, or can be twisted and wrenched and piled into odd shapes until they hint at being somehow perhaps on the verge or answering) a question that someone might conceivably want asked."

I worry about any type of bulk collection. If you collect everything, how do you know that you’ve got anything

The fewer data needed, the better the information. And an overload of information, that is, anything much beyond what is truly needed, leads to information blackout. It does not enrich, but impoverishes.

Like any decent researcher, I throw away 90% of my research.