There’s been a lot more attention for exciting new AI models and capabilities than for a deluge of state bills which have been rolling out to hamstring them in the name of avoiding “algorithmic discrimination.” I write for FIRE’s Newsdesk’s today about some signs of hope.

It’s an important clarification to popularism. You lean not on popular individual positions, but on positions within issues where you are trusted. Outside those issues, you lean on positions (which may poll so-so generally, but well with Rs) which build trust on the related issue.

To get at the point another way, check how universal background checks (for guns) tests and then how an ad where a Democrat espouses support for them tests. Voters like them in isolation but don’t trust Dems on the broad issue (guns) so they don’t get points for it.

Now put those policies together in an ad with any Democratic or Democratic/lefty-seeming candidate enthusiastically explaining those policies, and test that ad with the same voters.

Of course Trace was already doing the Yeoman’s work by the time I tweeted this. Some disappointing replies here from people who Just Don’t Get It.

I suppose the sudden elevation of somebody’s public profile is as good a time as any for some discourse about them. This does not have quite that feel — people are out for a scalp.

Personal feelings for the guy aside, the recent Cremieux stuff leaves a bad taste as, at a base level, it has felt like a retaliation campaign against the Times for using a disagreeable source in covering a popular politician, and against him for being party to that reporting.

I can imagine these exact dances performed by some twinkish zoomer for a Twitch stream or TikTok being used to make basically the same point.

‘Who let the dogs out’ really fell off. Don’t hear it around much anymore.

I am not sure I have seen a single prominent tech person that has moved right~ engage in introspection on that fact. They cannot resist flattering themselves with the idea their guiding principles have remained constant. To change your mind is just too human.

Very hard to understand in what sense the party has got meaningfully worse (absolute or relative basis) since this tweet. It’s ok to change your thinking! Self-reflection on that would be a lot more interesting than vaguely alluding to a NYC mayoral candidate and calling it a day

Notes: .1. You can say the poll is basically understood as a proxy for Pres approval by many Dems, but that speaks to point of engagement with heritage being shallow.2. The longer term Dem trend is probably education polarization, patriotic blue collar Dems leaving Clinton-Obama.

The back-to-back of these two posts on my tl is what gets me. This is what a patriotic event looks like to a driving share of the “very proud” cohort. That’s our level of engagement with the national heritage w/ liberal abstinence. Where’s the monster trucks?

Or rather “Trump has wielded civil rights to initiate the most significant assault on D.E.I. ” . Some reflection from these folks on how the same instrument that helped them bring it into the world is now stabbing it in the gut would definitely be the more interesting article

If you’re happy to use somebody’s changed name given they anglicized or shortened it, etc. but not when they are trans, you are mounting a protest. You’re free to do that, but in this particular instance (ie names) it is you who is breaking with the public status quo and not them.

Both the regulating state lawmakers and the masterminds of the moratorium are lacking in imagination. It’s very tough to regulate for unknown unknowns. I don’t agree with everything here but it at least gets at the implications of that problem.

Good post that backs up a lonely epistemic humility I have felt about the shape AI is going to take — one that has inclined me to both oppose all the counterproductive early AI regulatory proposals out there and the original 10 year moratorium on state regulations.

If the underlying legal question (which, btw, of course 1A protections apply to chatbots) doesn’t pique your interest, then the facts of the case — involving a ‘Daenerys Targaryen’ AI — surely will. Happy to have assisted on this, hope you give a read.

Does AI have First Amendment protections? Many have been looking for the canonical case(s) which will weigh that question, and a lower court has taken an early misguided stab at it in one possible candidate to be exactly that case. We filed to ensure their analysis doesn’t stand.

RT @TheFIREorg: And the First Amendment applies regardless of the tool people use to create their expression. Apart from carefully defined….

RT @LinkofSunshine: