8 Comments
User's avatar
Xian's avatar

Well said!! The work you deliver rarely surpasses the quality of the questions you ask.

Damien Faughnan's avatar

This makes a huge amount of sense. Thanks for making me wiser!

Yuji Li's avatar

Love this! So well said. Recommend read to folks in Strategy & Ops questioning “whether my job will disappear tomorrow?”

Shubha G's avatar

Brilliant! Great reminder that what is plausible is not equal to truth and business insight still requires contextual awareness.

Bharath's avatar

You are one of the people whose writing I actively look forward to. There’s depth, sharpness, and honesty in how you see things that really stands out.

nihal | deeptech decoded's avatar

One thing I’d add: first-hand experience. You can’t develop real discernment without having done the thing. Thank for sharing your perspective.

Joh L's avatar

As a data professional, I'm not seeing this. Right now everyone seems OK with data being the same black box as engineering.

Anh Le's avatar

You are right that software is more verifiable than data, and will be automated first.

But data also will soon be automated because all the context the "data curator" holds is ultimately digital and can be digested by AI.

Schema change, definition change: these are tracked in semantic layers. (I work on one of the, DataJunction)

Something in a remote corner of the company affecting the metric (a marketing campaign, an experiment) --is AI or human better at crawling through the mountains of context to find out what's relevant?

The convincing maybe the last piece to be automated. But even that would come soon once AI analysis is more reliable than human analysis