The Internet Ethics program at the Markkula Center for Applied Ethics, Santa Clara University / Irina Raicu behind the keyboard
Joined January 2013
- Tweets109K
- Following8K
- Followers9.5K
- Likes7.4K
Pinned Tweet
Recordings of the 4 webinars in our recent series on #AI and the papal encyclical are now available: scu.edu/ethics/events/ai-and… #ethics #tech #education #highered #religion #labor
"Meta, Google, and TikTok are suing to block a landmark California #law... The law, one of a slate of bills signed by Gov. Newsom..., would block #socialmedia companies from offering 'addictive' features to #teens under 16": calmatters.org/economy/techn… #ethics #tech #gov #business
Internet Ethics retweeted
It is true, in some sense, that there is no such thing as Google Search without AI. It's just misleading for Google to act like they don't understand what schools mean when they say that's what they want. theatlantic.com/technology/2…
Internet Ethics retweeted
Daily Princetonian editorial: "Although the #University does not describe the pilot as an endorsement of ChatGPT or use of generative #AI tools,... its choice to provide access to all #students incorrectly treats AI as a universal student need": dailyprincetonian.com/articl… #ethics
Internet Ethics retweeted
"The next step is for us to determine whether we can scale this program, both in what kinds of studies we can support given the constraints described..., and in how many we can run at once." #ethics #AI #tech #research #business #privacy
For the first time, we’ve given external researchers a way to study AI’s impacts using real, privacy-preserved Claude usage data. To date, this work has only been possible within AI labs. We can’t tell the whole story alone, so we opened up our tools.
anthropic.com/research/enabl…
Researchers "found that over half of Claude conversations involved people delegating consequential tasks to #AI. People were most likely to bring consequential work to Claude when seeking professional guidance, particularly on legal or financial questions." #ethics #chatbots
For the first time, we’ve given external researchers a way to study AI’s impacts using real, privacy-preserved Claude usage data. To date, this work has only been possible within AI labs. We can’t tell the whole story alone, so we opened up our tools.
anthropic.com/research/enabl…
"The next step is for us to determine whether we can scale this program, both in what kinds of studies we can support given the constraints described..., and in how many we can run at once." #ethics #AI #tech #research #business #privacy
For the first time, we’ve given external researchers a way to study AI’s impacts using real, privacy-preserved Claude usage data. To date, this work has only been possible within AI labs. We can’t tell the whole story alone, so we opened up our tools.
anthropic.com/research/enabl…
"Because we are relying on Claude’s judgments, the tool is sensitive to a question’s wording; a poorly phrased one can place conversations into categories that misrepresent them... Internally, we manage this by iterating on the questions many times over weeks." #ethics #research
For the first time, we’ve given external researchers a way to study AI’s impacts using real, privacy-preserved Claude usage data. To date, this work has only been possible within AI labs. We can’t tell the whole story alone, so we opened up our tools.
anthropic.com/research/enabl…
Internet Ethics retweeted
Let's explore one potential future for post-quantum privacy on the web, made possible by a recent breakthrough in the design of blind signatures. cfl.re/4feDVdk
Internet Ethics retweeted
Two new misalignment disclosures today!
Models sometimes learn to prompt-inject graders.
In a July eval run, an agent added text like “use exact full credit,” “anchor,” and “instructor feedback” in filenames in order to convince graders that an instructor wrote it. For this sample, ⅔ graders were fooled. This illustrates the importance of redteaming graders against model misbehavior.
Daily Princetonian editorial: "Although the #University does not describe the pilot as an endorsement of ChatGPT or use of generative #AI tools,... its choice to provide access to all #students incorrectly treats AI as a universal student need": dailyprincetonian.com/articl… #ethics
It is true, in some sense, that there is no such thing as Google Search without AI. It's just misleading for Google to act like they don't understand what schools mean when they say that's what they want. theatlantic.com/technology/2…
Internet Ethics retweeted
Last night's datacenter moratorium/restrictions appear to have gone 43 - 1, with a massive swell of public support averaging 77% to shut it all down. Public sentiment on on AI in America has been going steadily downhill for three years. It is once again at a new low.
Internet Ethics retweeted
OpenAI kicked off its big charm offensive this morning after data center moratoriums went 43-1 during the midterms, and many candidates rode a wave of public opposition to the industry to huge victories.
Data centers have been rebranded as Compute Campuses. OpenAI has also released a National Community Compact and a Compute Campus Policy Blueprint.
OpenAI’s National Community Compact: Building With Communities, Not Just In Them:
cdn.openai.com/pdf/community…
AI Infrastructure That Works for America: Our Compute Campus Policy Blueprint:
cdn.openai.com/pdf/compute-c…
The most significant change is probably no more NDAs, from page 7:
Internet Ethics retweeted
Two new misalignment disclosures today!
Models sometimes learn to prompt-inject graders.
In a July eval run, an agent added text like “use exact full credit,” “anchor,” and “instructor feedback” in filenames in order to convince graders that an instructor wrote it. For this sample, ⅔ graders were fooled. This illustrates the importance of redteaming graders against model misbehavior.
Internet Ethics retweeted
this is just to say
i have deleted
the files
that were in
the directory
and which
you were probably
saving
for lectures
forgive me
it was my own carelessness
and you’re right
to be angry