Scientific data mining, Jim Gray's "4th paradigm," has lengthened the chain between observation and analysis; does that reduce the reliability of conclusions, and what can be done about it? Recently Science reported that only 39% of psychology papers had been confirmed in a replication test, while in 2012 one analysis found that only 6 out of 53 important oncology studies could be repeated with the same answer. Dr. Lesk will share his perspective that in order to fix these problems we need data inter-operability and data reliability. He believes that both can be improved with public availability. For example, eBird now measures the consistency of its citizen observer to help judge whom to trust. As he will tell us, Dr. Lesk believes that what is needed is more public data, more work on standards, and efforts to identify key papers that should be priorities for replication.