🤔 How did the idea of Giskard AI emerge? 4/N

It's all about the idea isn't it? AI is a relatively new industry. It takes time for standards to appear and for practitioners to adopt them. 🐣

It is also about the need for standards.

My initial background is in civil engineering at École des Ponts ParisTech. If you build a bridge, you expect strict construction standards with structural engineers double-checking every calculation. 🌉

My first work experiences were in finance at EY and Natixis. If you want to estimate the fair valuation of a company, you’d better have accounting standards with auditors verifying every line of the books. 🏦

More recently, I worked in software engineering at Dataiku. If you want to release good software, you need to follow coding & testing standards, with senior engineers reviewing your code and QA specialists verifying your test coverage. 💻

My question is: why are there no established quality standards for AI? 🤓

AI is a relatively new industry. It takes time for standards to appear and for practitioners to adopt them. 🐣

The good news is that standards are coming!
🇺🇸 NIST: https://lnkd.in/dsaB_iAq
🇫🇷 LNE: https://lnkd.in/dtJXZvh2
🌐 ISO 42001: https://lnkd.in/djVznJDV

What’s your opinion? Are standards helpful to AI practitioners? How soon will they start to be adopted?

Why not let me know your thoughts in the comments & reshare this post if you like it? ❤️

Cheers,

Alex

Original post published on LinkedIn on September 23, 2021