Human Curation vs A.I.

robot hand using laptop keyboard

Human’s expertise is arguably mankind’s greatest achievement. Doctors, lawyers, software architects, financial analysts or any profession where an individual spends tens of thousands of hours to gain knowledge, produce experts that we trust to provide great advice. These experts provide us with specific, actionable, relevant advice on what to do to solve a particular problem.

It has been proposed that Artificial Intelligence (AI) will replace human expertise. Without weighing in on this too much in a blog format, we can state that in the short term, AI is not going to replace any form of real human expertise. Rather – what AI has become really good at over the last few years is Narrow Intelligence. When AI focuses on one narrow task it is referred to Narrow Intelligence. Examples of Narrow Intelligence include being able to transcribe human speech from a sound track, or ‘recognize’ a human face on a video (matching). Narrow Intelligence is sometimes contrasted with General Intelligence – which is what Humans have.

AI cannot do General Intelligence at the present time. Many contend it will never to it. Meredith Broussard recently released a great book on the subject called Artificial Unintelligence. I recommend it.

Curating a video playlist from a video library is an example of an activity that requires General Intelligence. Imagine a doctor that sees you at her office and to fix a wound that you received while trimming your bushes. The doctor stitches you up, provides you with a prescription to a drug and sends you on your way. Then the doctor picks out 2 or 3 videos that you can watch at home that describe care of the wound and more information about the drug. These come to you in a playlist that you can watch from your computer. This is an act of curation.

Curation is the act of selecting content of the greatest interest or relevance to a person or group. This is something that requires a human expert. But how can the computer help here? What portion of the problem can AI help solve? What can be automated?

Creating searchable timecode-based metadata to make video content more discoverable a is referred to as Semantic Enrichment. Semantic Enrichment is the act of adding metadata to content to make it more usable and therefore more valuable. This IS something that AI can help with.

While humans are better at knowing personal or group preferences, priorities, and intentions, computers are better at rote behaviors such as finding similarities, recognizing objects, and transcribing. Leveraging the relative strengths of each (people/curation and machines/enrichment) provides the greatest benefit for the consumers of content.

Common Caches logoInferno logo

Jolokia and Common Caches have integrated their video-focused offerings, Inferno, and Media Share, respectively, to provide the creators and organizers of content with a broad set of tools and capabilities to curate, enrichment, and deliver content to audiences anywhere and of any size. This combination of automation and human curation is the perfect match of technology and human expertise. We urge you to give it a test drive.