winograd, terry
see uber
---
terry Winograd check YT
Stanford D school?
What's up with AI?
February 20, 2026
The current boom in AI has been accompanied by tremendous hype, both negative and positive. My goal is to go beneath this surface and provide a better understanding of what AI systems are actually doing, and what concerns we should have about where they are going. I am neither a doomer nor a booster. The very real problems created by AI today and in the foreseeable future need to be approached by considering the ways we (as a society) choose to apply it and the ways in which it can fit into our world. The bottom line is that we need to approach all of the issues with the recognition that large language models and similar systems have no real understanding or intention, even though they often promote the illusion that they do. They can still be useful in many contexts, as long as we recognize the need for humans to always provide the foundation of care.
student no work w/ peer on homework problem, just chatgpt to get rt answer
eliza chatbot pretended to be psychiatrist
wd you mind leaving office bc this is getting personal
“always” all men are...--> can you give specific example?
“mico” someone who truly listens MS; Bob melissa french stnfd commun dept disaster bc not ai; sycophancy to boost engagement --> encourage suicide
no way to control ai once you put info in?
surveillance
admin subpoena soc med corps; gigabyte to ID the ---- terrorists
SDI will make nuke missles impotent
david parnas cs for social respons
robo warrior/ cop not as scary as cmd and control decisions
josegh weizencoppf bk
tchr: claude boys do what it says nihilism
heidigger on tech
56:06
Yeah. And the answer is, people always are fascinated by science fiction things and by where things might go and to the extreme.
56:13
I mean, that's a really interesting thing to talk about. So it's taken seriously, I think, by people who aren't looking at the current things
56:21
as much as they are projecting. I mean, you look at things like effective altruism and the futurists and so on.
56:27
They're saying, well, we need to really do the right thing for these trillions of people who are going to exist a million years from now,
56:32
and so on and so on. That can be fun. It's just not what's going on today.
56:38
Now, what about what do people like in HCI and other fields have to offer? And I think it's hard to pin that down because it's not like,
56:45
well, here's a technology to use instead. It's like, well, let's question how we're using technology in general.
inequality
ezier to make sure tianamin sq not known
---
Diyi Yang via pcd-seminar <pcd-seminar@lists.stanford.edu>
Mon, Feb 16, 6:30 AM
to hci-research@lists.stanford.edu, pcd-seminar, cs-seminars@lists.stanford.edu, cs547-win2526-students@lists.stanford.edu, cs547-win2526-staff@lists.stanford.edu, nlp-group
What's up with AI?
Terry Winograd, Stanford
Human-Computer Interaction Seminar (Seminar on People, Computers, and Design)
Feb 20, 2026, 11:30am-12:30pm PT
Gates B1
https://hci.stanford.edu/seminar
Open to the public · Previous talks are online at https://hci.st/547youtube
---