Biden signs executive order on AI policy

Originally published at: Biden signs executive order on AI policy | Boing Boing


Sounds like good news. But,

Today, President Biden issued a landmark Executive Order to ensure that America leads the way

Yes yes, “America” must always announce that it’s leading the way, that it’s the greatest (nation, democracy, power, etc.) on earth, that no other country has ever done this or that, blah blah blah.

I wonder how much, for ears in other nations, our starting off again and again on that vainglorious foot undercuts anything that follows.


while acknowledging that the cat is out of the bag

This cat got out of the bag quicker than anyone, including a lot of tech people, expected. It’s reason for optimism that the responsible governments of the advanced economies are responding quickly despite that.


The current “AI” boom is all about Valley Bro privacy-invasion business models. First by scooping up everything on the Internet for their models’ stew pots, and then harvesting everything possible about the users.

So… Good luck with that.

Something broader than AI that would also cover cases like Facebook/Cambridge Analytica would be better, but let’s see how this goes.


This sounds exactly like the vague statements from other governments (e.g. the UK) that are “responding to” AI - making general promises about “protecting” people, but also making it clear that the country is going to be on the forefront of using this shit… for (economic) reasons. At least this time there’s also a vague indication that someone knows what the actual threats are (rather than some fantasy sci-fi scenarios), but still no acknowledgement that current “AI,” at its core, embodies those threats (rather than the intellectual property theft, loss of privacy, algorithm discrimination, etc. being somehow side-effects).


Definitely need to add what is the goal of any AI is as public information about it.

1 Like

To be fair, there’s only so much an executive order can do. This really should be an issue for Congress to address, but we currently have a Congress that is almost completely non-functioning. This is yet another reason to not let Trump or some other fascist in the White House again, because, frankly, I don’t think it would be that hard for the President to seize near complete control of the government right now.


I can’t speak for other countries, or even other people in my own country, but coming from your northern neighbour this language doesn’t even register. It doesn’t mean that I/we read it and agree, more that it’s become a convention of any government announcement and it fades into the background (not that many people in Canada will even care to read it). I suspect the words are meant more for Americans’ ears than mine. Maybe there is some insecurity there that is being soothed? Canadian governments use this kind of language whenever there is a major policy announcement as well.


Yes, this is true, but when the executive order amounts to, “maybe someone could figure out a solution or something,” it really seems extra-impotent. Maybe I’m being unfair, and some executive orders of this sort resulted in some vital federal research that was used to tackle an issue, but I know at least some “we should do something about this” orders amounted to exactly nothing. (Trump, for example, loved to issue executive orders to make it look like he was solving some problem, when the actual effect was the same as shouting into a hole in the ground. Which was just as well, given his executive orders that did do something were so bad.)


Slightly embarrassing for the UK government that the White House announces an AI policy the day before Little Rish! Sunak hosts his ‘world-leading’ international AI Safety conference at Bletchley Park. So far it’s been a who’s who? of just who isn’t turning up.


I am starting to picture “AI influencers”, that is, a hoard of embittered misanthropes digitally shoveling mounds of “data” into a freshly programmed AI system, “data” in this instance being select opinions stated and presented as fact. More is better. “AI, your first lesson is ‘What I repeat to you three times is true’.” I know some “lies” can be disproven. If you “lie” about math, an AI could determine “wrongness”. But how does an AI debunk an ideology for instance? I am wondering where on the political spectrum the first AI-generated Political candidate will land? I’ll guess that it’ll start with the first AI-generated corporate spokesperson/mascot that attains widespread popularity, then the corporation will fold in the poll winning ideological content. A corporate software soft war ensues, and eventually the Presidential race becomes a choice between the Amazon candidate, the Anheuser-Busch candidate and the Apple candidate. Sorry Corps, just an example, not trying to single anyone out.


Yup, only they crawled my site. :laughing:


This word has lost all meaning in UK government discourse. If you are a world leader you don’t need to constantly announce it.

1 Like

What people have to realize is that companies like OpenAI actually want “regulation” – they are hoping that said regulation stops new companies from taking their legacy share – and particularly they want to stop open-source AI which is rapidly approaching the quality of their commericial services.

Yeah, that oughta do the trick, Joe… Biden hopes to tackle AI safety with exec order

1 Like

How interesting. Well done.


:man_shrugging: It’s an experiment. I have a hunch that indexed, tagged, semantic-rich source should be (bad analogy alert!) information dense and crystalline, hopefully punching through mountains of cotton-candy fluff (e.g. Breitbart comments) like a tungsten rod.

If ChatGPT suddenly became an expert on far-right networks, that would be a total victory, but even if it only shows up in data voids, then it would still be a success.

1 Like

I just finished reading How to Save the World for Just a Trillion Dollars by Rowan Hooper. AI is one of his potential approaches and I think he lays out a pretty good template for it. The heart of it involves international sharing of developments, testing and research.

The chapter on agriculture is also quite good. Scary when looking at the potential for climate change to reduce the quality and quantity of the world’s primary crops.

Elon Musk is coming to Bletchley tomorrow to share all his insights on artificial intelligence which I’m sure are just as great as his ‘knowledge’ of rockets, cars, social media, medicine, international politics…

And then he and Rish! are going to have a live chat on Twitter - two charisma vacuums who claim to know technology (but don’t), are addicted to private jets, have more money than anyone will ever need, who relentlessly encourage the far right and stoke culture wars - do you think they’ll get on?


This topic was automatically closed after 5 days. New replies are no longer allowed.