If you're headed to #DragonCon2024 don't miss Meredith Rose!

On Sept. 27, Lawfare and @GtownTechLaw with support from @OmidyarNetwork, hosted "AI on Trial: Liability in the AI Ecosystem," a conference on the complex landscape of AI liability. It included a discussion of legal primers that can be found below. youtu.be/zpQwDcRx9IU

Excited to share this at last! 🎉 This paper on AI agents pulls together findings from a fascinating workshop we held in May. We hope it can be a helpful resource for anyone who is hearing all the excitement in tech circles about "agents" and isn't sure what to make of it. 🧵:

Happening soon! Tune in for what is sure to be a great discussion and overview about IP issues related to generative AI.

📢A major coalition is calling on Congress for action on AI today. We're talking non-profits, universities, AI safety groups, consumer advocates -- even the tech industry itself. Our message to Congress: authorize the US AI Safety Institute and lead on AI testing and

This is exciting news! It is tremendously important to have independent nonprofit organizations with deep technical understanding for there to be a true AI accountability ecosystem. Companies shouldn't have a monopoly on understanding the systems they build.

Please go out and vote today! It took me less than 10 minutes, it is a lovely fall day, and it is your chance to participate in our democracy! Go #Vote2024!

The limited Constitutional scope of copyright helped create the Internet, says Nicholas P. Garcia Public Knowledge. It incentivizes creators to make new works, ensuring a *broad public domain* - the corpus of Internet content. 🤖 Full video/audio here 📺 netcaucus.org/event/intellec… 5/

This is purely anecdotal, but I think a lot of very online Democrats need to really grapple with this.

Tomorrow 1 pm ET: I'm looking forward to discussing whether AI is woke with Vincent Conitzer, Nicholas P. Garcia & Kenneth L. Marcus on a Federalist Society webinar: fedsoc.org/events/is-ai-w… #ArtificialIntelligence

We are proud to unveil a new vision for free expression and content mod by Policy Director Lisa Macpherson and Policy Analyst Morgan Wilsmann. Our goal is to review big media changes and describe how policymakers should think about content mod. Read: publicknowledge.org/ContentModPrim…

But despite the potential for AI to radically change our economies for good and for ill, the main policy energy seems to go into *accelerating* AI. More attention should be given to safety. Completely absent is a focus on *directing* AI deployment towards good outcomes. (4/8)

Looking at the replies on Yang's post shows what a cesspit this site has become and how bleak things are for our republic.

BEAD is a once-in-a-generation investment. LEOs are a promising tech with real use cases — esp where fiber isn’t economically / physically feasible — but we should continue to let states choose the tech that works best for them, not force more LEO uptake. publicknowledge.org/slowing-bead-t…

Today, the House of Representatives passed the “Tools To Address Known Exploitation by Immobilizing Technological Deepfakes on Websites and Networks Act" #TAKEITDOWNAct—a bill designed to address serious concerns in the digital era, but that could also harm free speech & privacy.

Teddy Roosevelt's new nationalism speech, 1910

3/ We need smart, balanced legislation that protects commerce, privacy & dignity without sacrificing free expression. Read the full breakdown and recommendations in the latest from Senior Policy Counsel Nicholas P. Garcia: publicknowledge.org/digital-replic…

Mandating the licensing of pretraining data sounds clever until you realize that even if a trillion token model were to exceed the yearly profits of google, the most valuable data would at best get a few dollars and 99% would make a fraction of a cent More imagination needed

I'm not a gambling man, so people will have to settle for my hopes and fears instead of wagers and predictions about tomorrow's AI action plan.