111: The Ethical AI Layer Cake
from Fresh Fusion Hosted by Jared White
In this episode, I dive deep into how I might be willing to acknowledge an ethical approach to using generative AI tools, and why virtually all tools in widespread use today fall wildly short in the morality department—which makes the embrace of these tools by corporations incredibly disappointing. But can we thread this needle? Is there a viable path forward? Let’s discuss. Links & Show Notes Sign Up for Intuitive+ to listen to Fresh Fusion+ Leave a review on Apple Podcasts Subscribe to Creator Class (a Newsletter) Follow Jared on Mastodon & Pixelfed See the comments on this very episode, and add your own! Shared Links: The Subversive Hyperlink - Jim Nielsen’s Blog RMTransit on PeerTube — 698 videos on public transit & urbanism! Reece Martin’s Blog Kickoff Toots: I don’t think many folks who primarily spend time on Mastodon as their social network of choice realize how absolutely bonkers it is elsewhere… Just so we’re super-duper-crystal-clear here, there is no AI, none whatsoever, that anyone anywhere has invented which actually thinks…actually reasons… Crazy Stat: According to a 2021 Georgetown University poll of gun owners in the US, 24.6 million persons have an AR-15 or a comparable firearm in their possession. –Wikipedia
Embed this episode
What this episode covers
How I *might* be willing to acknowledge an ethical approach to using generative AI tools, and why virtually all tools in widespread use today fall *wildly* short in the morality department—which makes the embrace of these tools by corporations incredibly disappointing. But can we thread this needle? Is there a viable path forward? Let’s discuss.
Chapters
- 0:00 Introduction
- 2:37 The Subversive Hyperlink
- 4:27 RMTransit is now on PeerTube!
- 6:47 So I'm probably not a likely person to get excited about generative AI…
- 14:21 An attempt at an objective ethical AI framework
- 19:02 I'm not an AI "alarmist" or "doomer"
- 20:27 Ethical AI Layer Cake
- 21:17 Layer #1: 100% openness
- 22:58 Layer #2: Full transparency around data sources
- 27:34 Layer #3: Sources must be 100% opt-in
- 29:11 Layer #4: Tools should be narrowly purposeful
- 31:48 Layer #5: All generative AI output should be tagged as such
- 34:07 Layer #6: Users must choose to opt-in to using generative AI
- 36:20 Optional Layer #7: We need environmentally responsible AI
- 37:42 Summary of the ethical AI layer cake
- 40:33 Unethical generative AI is not "inevitable"
- 42:54 Why are we not seeing proper ethical AI frameworks promoted by AI vendors?
- 44:35 Today's computing hellscape and outrageous punditry
- 46:26 Wrapping up
- 50:00 Outro
No transcript yet.