Hard to see how folks can keep saying AI isn’t a part of the Google Search antitrust case when Google is selling search grounding as an AI service:
Posts by All Joe Jeromes Are Belong to Us
Of course, Google will fight this tooth and nail, and probably have AI draft 25% of their talking points about data access remedies. Full article is here:
"So the proposals from the DOJ that match DDG’s suggestion of API access are much more interesting (and probably better) overall… It creates a scenario for increased competition without a corresponding harm to users or to other competitors.”
Couldn’t put it better myself!
Taking a World Series programming break to notice that @techdirt has a piece out on what to do about Google’s search monopoly: "The most reasonable suggestion seemed to be @DuckDuckGo main suggestion."
Meta AI’s “open access” includes usage restrictions that prohibit the use of Meta’s model to train rival LLMs and "the need to acquire a license once a business surpasses 700 million users,” note @danielahanley and @maxvonthun
So @DuckDuckGo wants signal to startups, potential partners, and the larger tech community that our door is open to anyone who wants to work with us on a more private Internet:
Meta @OversightBoard recommends changing community guidelines to stop the sharing of private residential information when it is considered 'publicly available’ in 2021.
Three years later:
“Not covering the extraordinary story of Concord reveals a deep failure in the mainstream press, and one born of a silly prejudice against a medium that towers over the films and television that such outlets are willing to understand.”
Truly fail to understand how this helps users. Do get that this is another way for Meta to try to boost engagement.
Google broke the law. Now argues fixing that will destroy capitalism.
I look forward to seeing what Google believes is a reasonable restriction on its business. I guess we’ll all find out in December! G’night!
Considering the other adtech competition case, I think you take Google’s claims about online advertising with a grain of salt.
Again, parallels to the breakup of the phone company. This is effectively platform hostage taking, and a divorce between OS/browser and search certainly doesn’t break the products.
“America’s technological leadership” depends on protecting a monopolist’s ability to dominate in AI. It’s a compelling rhetorical argument, but it’s functionally little different from previous arguments that breaking up the phone company would hurt national security.
A few quick thoughts:
(1) Sharing click and query data is already the law of the land in the EU, which has stronger privacy rules than the U.S. anyway.
(2) Google’s highlighted argument is a hedge. DuckDuckGo argues the opposite, that this increases incentives to differentiate.
And @googlepubpolicy is out with a response to the DOJ’s remedies framework (and is engaging in a bit a publishing time travel):
Anybody else just refreshing, waiting for the DOJ’s remedy framework for Google’s anticompetitive search practices?
“Privacy is valuable not because it empowers us to exercise control over our information, but because it protects against the creation of such information in the first place.”
Oh dear. First Amendment Twitter is gonna be triggered.
AI and copyright is fraught, but I take some issue with statements like this from a company built on UGC.
Meta just announced the Hyperscape app at #MetaConnect2024, which allows people to explore detailed maps of spaces. Just your gentle reminder that detailed spatial mapping raises some interesting privacy, property and safety considerations:
Earlier this year, I joined my old boss @MichelleDeMooy to talk about the Youths and tech policy for @McCourtSchool’s Asterisk podcast:
We can open up Google without opening up user data. Licensing Google’s search results does not involve accessing any user data. This remedy will not invade users' privacy! 5/5
Search engines could more easily differentiate on things like privacy and UX/UI. Imagine universe of search experiences, such as features that allow users to tweak ranking algorithms, features that bring more transparency to ranking algorithms, and other AI capabilities. 4/x
The best way to level the playing field is for Google to provide access to search results via real-time APIs on FRAND terms. That means any query that could go in a search, a competitor could access the same search results. @KagiHQ and @TC_4KBW have also suggested this. 3/x
Screw you, NYT Wordle Bot.
California’s #AB3048 would require your web browser offer a simple opt-out signal like @globalprivctrl, and as with DNT, industry groups seem to be insisting people declare their privacy preferences over and over and over again.
Can a lawyer (or a data broker stooge) explain to me what the difference is between the “do” and “don’t”?