Too often, AI solutions are built without a deep understanding of the human impact, leading to unintended consequences. This program by @StanfordHAI and @stanforddschool aims to nurture a community of informed civil society leaders to address this gap: https://
stanford.io/3SA6kPG
ETHICS
-
Stanford Program Addresses Human Impact of AI Solutions
By
–
-
Self-Improving AI and the Darwin-Gödel Machine Deep Dive
By
–
AI that can improve itself: A deep dive into self-improving AI and the Darwin-Gödel Machine. https://
richardcsuwandi.github.io/blog/2025/dgm/ Excellent blog post by @richardcsuwandi reviewing the Darwin Gödel Machine (DGM) and future implications. -
LLMs lack memory consolidation like coworkers with anterograde amnesia
By
–
Agree that this is an important capability hole right now (I saw you push back on it a few times in the pod and I also didn't find the answers too satisfying). I like to talk explain it as LLMs are a bit like a coworker with Anterograde amnesia – they don't consolidate or build
-
Regulatory risks for startups in restrictive jurisdictions
By
–
It's not just the culture. If you have laws on books that will get you caned for chewing a gum, then there are probably laws that could get you in pretty big trouble for something that your startup does that is not entirely under your control.
-
Fragility of Assumed Technological Solutions and Economic Risks
By
–
This sometimes worries me, because things we take for granted as being Utterly Solved are much more fragile than they appear, and treating them as harmless places to extract economic rent (or similar) risks far faster breakage than people would expect.
-
Innovation Barriers in Regulated Industries vs Low-Barrier Sectors
By
–
But it only works for lightly regulated/low barrier to entry industries. Try being innovative in medicine. Or even any kind of manufacturing.
-
Software Engineer’s Website Forces Kenyan Government Action
By
–
Finally. Free.
— Erik Hersman (@whiteafrican) 3 juin 2025
This little lady with a website sure scared the Kenyan government…@rtunguru thanks for showing what an impact software engineers can have! https://t.co/cuVsvWKsneFinally. Free. This little lady with a website sure scared the Kenyan government… @rtunguru thanks for showing what an impact software engineers can have!
-
Test Examples Extractability and Privacy in Saturated AI Models
By
–
– hrunning these experiments in a clean setting with perfectly deduplicated texts tells us a lot about privacy: – once capacity is sufficiently saturated, the **test examples** are slightly more extractable than the training examples — maybe extraction is a bit of a myth?
– the -
Text Communication Challenges and Professional Workplace Concerns
By
–
It's so tough to explain things in texts sometimes, and you try to be nice but still sound rude to others. I hope I won't get fired one day because of this.
-
Government Liability vs Tech Innovation Responsibility
By
–
In this case, if the *government* is negligent or liable (not Tech Bros) that changes everything.