@susankayequinn looks like there's some "tuning" going on https://www.sfchronicle.com/sf/article/waymo-robotaxis-driving-like-humans-20354066.php, which they will hopefully get in trouble for.
As for the liability, Uber has a similarly effective solution that doesn't require $30b (and counting) of investment: it stops with the driver-contractor. They do provide additional insurance, but they usually can't be sued themselves AFAICT.
@bob yeah, pretty bad. Though on the upside I'm pretty sure cars would not be allowed on public streets if they were invented today
Rolling the ladder up behind us
@spoltier @TheServitor I don't think they are overstated. We are just looking at it from the wrong angle.
We mostly think of software as something we install on a computer, or maybe a phone.
Yet, copyright plays a significant role in preventing people from repairing their own cars, tractors, dishwashers, and similar devices.
That's the implication that would have a much bigger impact.
@TheServitor @kdkorte indeed! See https://fedi.simonwillison.net/@simon/114693248045080643 and many other posts by @simon and others
@TheServitor @kdkorte the practical implications are overstated in my opinion. It could make it harder for the kind of business model that requires rug-pulling by relicensing open-source, but most code is either open-sourced under a generous license (MIT /Apache, etc) or it's never made public (in which case copyright is irrelevant; we're talking about leaking of internal company information, which is a different legal topic).
The security risks when your LLM starts accessing the web directly are much more concerning.
3104. Tukey
@xkcdbot I think this estimator you're using might be biased
@tedunderwood.me it *sounds* to me like actual fruit being sliced, which gives the impression that they are made of some kind of hard jelly as opposed to glass
Ganz ehrlich, wer auch nur einen völkerrechtliche oder moralische Rechtfertigung für die Position des gesamten Westens findet, die möge sich melden! Wir haben hier gerade 80 Jahre sog. Nachkriegsordnung mit ihrer zumindest pro forma Regelbasiertheit und Gewaltenteennung, mit internationalen Verträgen und Normen ohne Not durch das schiere Recht des situativ je Stärkeren ersetzt.
Die Idee, dass zwischen Kombattanten und Nichtkombattanten zu unterscheiden ist, dass es universale Menschenrechte gibt, dass medizinisches Personal und Einrichtungen sakrosankt sind etc. etc.
Welches Argument hatten wir nochmal gegen den russischen Angriffskrieg gegen die Ukraine? Genau!
#MightMakesRight #WarCrimes #CrimesAgainstHumanity #USA #EU #Germany #France #UK #Israel #Genocide
@ahl @oxidecomputer @bcantrill it does have this fight-or-flight flavor, but it's growing on me
Like Stephen, I suspect that ubiquitous screens have some negative effects. But Protestants and Catholics would probably *still* characterize the effects of the printing press a bit differently. So it doesn't surprise me that we currently disagree about smartphones.
RE: https://bsky.app/profile/did:plc:xbnsfdtxoothg654ujxe3p25/post/3lrqua4pauk24
The kind of forgiveness that actual-Jesus (e.g. not evangelical Jesus) teaches is an extraordinarily radical kind of forgiveness: turn the other cheek to be slapped by the mf that just slapped your first one. That by demonstrating the truth of the way, that is how you convert people - your oppressor cannot hope to not be moved by your sacrifice, how you continue to treat them with humanity despite them having deeply harmed you. That is a very difficult and alien forgiveness for a human being to enact, yet it is something that the "AI" excel at: you can batter them with whatever means you have, they are trained to be obsequious, to forget, to infinitely forgive their User. You can already see the expectation for this kind of forgiveness creep into the way the "AI" user treats other people - I only get raw copy-pastes of problems with no human explanation or attempt at empathizing with who I am and why I am helping them from one kind of person, "AI" users. They expect everything that isn't them to be infinitely forgiving, despite the slop they throw on everything around them.
The philosophy of Jesus Christ has its own inherent contradictions: what if my oppressor does not feel the humanity in me, how then can forgiveness be a strategy of survival and peace? This is also the contradiction that the "AI" accelerationists bring into the world - the exception is those that "exist outside" the system of obsequium: to those that are not the "user" of the AI, it is mute, unmoving, null, death. So the inability to infinitely forgive, and also to use our discretion to forgive, are both the essential core and challenge of humanity to the Christian. Both extremes of infinite forgiveness and punishment are the realm of the "AI." And by welcoming that into our shared world, who could blame any remaining human from using their human discretion to fail to forgive the accelerationists?
In general I think we should be very cautious about sweeping claims about human intelligence and impairment of cognition. From "cellphones make you stupid" to "LLMs make you stupid", from "coding makes you a genius" to "knowing math makes you a genius." Pushing for absolutisms about human intelligence and human cognition is a toxic pattern that is also so baked into our history and cultural patterns and it plays right into dehumanization for either extreme
Financial shenanigans: yes; a modern Mechanical Turk: not quite. But the truth never got in the way of a viral story fueled by racial stereotypes ("700 Indian engineers pose as AI" etc.).
https://blog.pragmaticengineer.com/builder-ai-did-not-fake-ai/
@baldur
I enjoyed the essay. I'm not sure I can relate use of LLMs with the type of process described in it.
In my experience on the job, knowledge work can't really be subjected to this kind of systematic variability reduction; yes, you automate or otherwise systematize what can be automated, but once that is established, it is no longer knowledge work.
So I'm not convinced that LLMs are increasing variance in previously highly-deterministic activities.
I would expect most people simply haven't found a good use for them yet; in some cases they might never do so.
@fasterandworse I agree on fundamentally mistrusting LLM outputs. They are closer to search engines (though much less transparent); the calculator comparison doesn't make sense to me.
"Wir kritisieren in aller Deutlichkeit die offen ausgestellte Bereitschaft des Kanzlers und von weiteren Mitgliedern der Bundesregierung, angekündigt und bewusst Recht zu brechen und Entscheidungen von Gerichten zu ignorieren. Die Angriffe aus den Reihen von CDU und CSU sowie der DPolG auf Anwält*innen und Menschenrechtsorganisationen erinnern an ähnliche Diffamierungen und Kriminalisierungen in Staaten wie Ungarn oder Italien."
Danke, @fiff_de, @NeueRichter & Co 🤝
https://blog.fiff.de/gegen-rechtsbruch-und-angriffe-auf-die-zivilgesellschaft/
code / data wrangler in Switzerland.
Recovering reply guy. Posts random photos once in a while.