It's strange to me, as it appears that in the US it apparently doesn't matter if any charges have already been filed. In my country a pardon can only be given to someone, who was already sentenced - there was even some controversy when a couple of years ago the president gave out a pardon before the legal process concluded. As a PhD in law he should have known this made no sense.
You could certainly argue US pardons are supposed to work the same way too, it's just no one has standing to challenge the blanket pardons, but I don't think anyone has tried recently. And the current regime is happy to keep such things around so they can use it on the way out the door.
Correct - it's even weirder, pardons can be given for a categories or even just a blanket pardon for any crimes that might have been committed. There aren't a lot of limits on it.
USA won't extradite their war criminals. Look how
desperately Trump tries to fight down The Hague
criminal court. They have a LOT of dirt to cover
up here. I mean, just take the bombing of a school
with girls in Iran - is that not a war crime? And
that's just one of many more examples one could
give; Trump denying data to Ukrainians via Starlink
means he supports Putin's genocide.
Iran is actually fair game since it doesn't want to be part of the ICC. The same was true for Iraq and Afghanistan. There are no rights without obligations.
However, this isn't the case with Palestine, which is why Netanyahu has problems - and obviously Putin as well.
Israel and the US are also not part of the ICC. Yet they still call out "violations" of international law all the time. The ICC is just a colonialist tool used to beat down any trend towards sovereignty.
Kushner may not, but the excerpt indicates that Kushner’s lawyers only performed actions with approval of the president and in the interests of the United States.
So… the President is in pretty hot water if any senators decide to ask a few questions here.
Ah yes ... looks at calendar ... it's a Friday ... whooo boy, the President is in some pretty hot water today! Like yesterday, and the day before that, and the day before that, and all the tomorrows until he's not the president any more.
Seriously, there is absolutely nothing the current US president could do that could constitute a "hot water" situation for him give the willingness of the Republican party and the SCOTUS to support him unconditionally.
Clear, extrajudicial funding outside of the law seems like one of those things? Iran/Contra comes to mind as a point of comparison but at least that had some tacit bureaucratic process to it.
I think there's a question of semantics here. RL is way more expensive in terms of compute, as you have to generate a significant amount of output before you can judge whether it's good or bad. So compute is one angle, but another is how much impact RL has on the weights vs pre-training
They no doubt have more expensive/powerful models internally, but smaller models seem to catch up fast. So I'm not sure it's about capabilities, but more the willingness and budget to conduct a huge search.
Obviously the more intelligent the model, the smaller/more directed the search is. But they spoke about huge numbers of agents working on Navier-Stokes for example (I think it cost >$10m).
All the methods and data are not public. We don't know what unpublished methods they're using. You can get most of the pre-training data publicly but they've probably spent a ton of money curating it and are now doing things like buying rare books. The RL training data is all (/mostly) proprietary though, and that's the real secret sauce part.
All the RL data are exactly public. There are huge amount of distilled data freely available, and that amount is more than enough to train a ~10T model.
There are 'transfer stations' and that's how exactly I use GPT and Claude in China. OpenAI and Anthropic do not sell in China, so we use their AI with a much lower price like 1% of the official API price. The largest transfer stations have TBs of traffic every day, and the traffic is eventually possessed by the open source community.
Subscription engineering is a deep field. Neither OpenAI nor Anthropic have any technical advantage in this field.
I think a big issue is that the companies pushing AI want it to replace people entirely. So the software around it is designed with that mindset.
I'd really like something that works more like pair programming. Where you share an editor session with the AI, and it's much more interactive. eg. It says "I'm thinking about doing X here, what do you think?". Then you give a response and it continues. Or you can see it doing something silly and immediately stop it and correct something either with a prompt or by manually editing yourself, before letting it rip again. Or just ask it "why are you doing it that way?". Maybe with some control over the speed it's going, for when you feel confident about what it's doing.
Claude has made some positive changes over time where it asks you more when there are different approaches it can take. But I feel like I'm not being brought along with my mental model as much as I would like. A lot of the time it spits out a whole pile of code and then I have to go and build up the mental model after the fact, and correct a lot of what it's done.
The current approach is good a lot of the time though, when you're not really changing anything architectural and just want it to bash out code while you do somehthing else.
It's the exact same training process for both of your examples. I don't really see how you can claim books are not replicated, but that output from other LLMs is.
Process is the same, but intent is not. One intent is to extract information from the book for better general eloquence and overall awareness - not for replicating the book itself (ability to recall verbatim fragments is a side effect, not the goal). Another intent is to replicate the behavior, carry it over using training.
Again - IANAL - but in my understanding (and I spend some time reading on this), the legal concept of fair use is all about the intent how copyrighted material is used. It's all copying or distribution, but law does make distinction about what and why.
They have no choice but to train their own model to try and survive. They're paying API pricing for the top tier models but competing against subsidized subscriptions.
