building the Q3 2026 edition of our AI fluency assessment at aisa.to and one pattern keeps surprising me — Safety is still the weakest dimension at 45/100 across all assessments.

people are getting better at prompting, better at picking tools, even better at multi-tool workflows. but almost nobody has a systematic approach to checking whether AI output is actually correct before using it.

we're adding new evaluation criteria around verification practices for Q3. the question isn't "do you use AI" anymore — it's "do you know when to trust it and when to double-check."

if you want to see where you stand before Q3 drops in october: aisa.to

#buildinpublic #AISA

Ozan

Activity

Products

Recently upvoted

Posts