~/talk

Is Your AI Truly Open ? Cutting Through the Hype

on YouTube 37m
Talkat DevConf.CZ 2026 · on Friday, 2026-06-19· at 11:00· D105 (capacity 300)· [open-source, large-language-models]

speaker

name:Alfonso Cancellara
links:none on file

resources

video:https://youtu.be/LbZp3LGmCx8
slides:none on file

summary

Defines concrete checks: use the Foundation Model Transparency Index, Linux Foundation Openness Framework and SPDX 3.x model sheet to verify released artifacts (training code, data provenance, model parameters) rather than trusting labels. Points out models with Apache‑licensed weights like GPT‑OSS still score low because they omit datasets and training recipes, that frontier models have ~10 trillion weights, and that many claimed “open” releases are open weights vs open source and amount to openwashing. Practical fixes include indemnity, dataset documentation and index scoring.

AI-written summary · about

chapters (9)

  1. 0:00Introduction
  2. 2:09AI definition today
  3. 8:55Open source fundamentals
  4. 11:42Defining open source AI
  5. 17:00Open datasets limitations
  6. 22:57Open weights vs open source AI
  7. 26:32Open washing in AI
  8. 29:00Call to action
  9. 29:42Q&A

related talks

~/search
type to search 6,249 conferences, 138,987 speakers and 185,172 talks
~/help
cmdctrl k
search every conference, speaker and series
/
focus the filter search
h
home
c
conferences
s
speakers
e
series
t
light / dark — the status bar button does the same
?
this help
esc
close this, then clear every filter — or just the search, while you are typing in it

Ignored while you are typing. Pages without a filter rail have no search to focus.