A decent chunk of the work a student does during a PhD is stuff that their advisor would have done in much less time, but are intentionally delegating to the student to allow them to learn.
So, I don't think AI changes the way expertise is earned. It might change the cost tradeoffs depending on where AI costs eventually settle.
Ok, but that’s a different thing altogether. I personally expect a ban, that doesn’t mean it will be very effective, but that does mean it will have a meaningful impact (such as not being able to offer open weight models as a service)
I don't believe this is raising the bar for minting fresh PhDs.
We need to maintain perspective here, a PhD is essentially work done by a researcher at the least experienced, least skilled point of their career. Their primary goal is to demonstrate that they are capable of contributing to research.
They are not competing against AI to publish a counterexample to a known conjecture.
They should sell an add-on to your subscription so that if Claude is down it routes the queries to OpenAI. Then OpenAI can do the reverse for their subscriptions, and if both are down, Grok will invest your portfolio into Nvidia.
The true intention of all of these seems to mostly be to give the government justification to pick winners and losers. Anyone that bends the knee gets their carveout or special exception, anyone that doesn't gets hit with all sorts of restrictions on top of the randomly applied tariffs.
It seems the way to get Americans to embrace a centrally planned economy was to combine unashamed xenophobia with blatant corruption while calling it capitalism.
>The true intention of all of these seems to mostly be to give the government justification to pick winners and losers.
AlwaysHasBeen.jpg
You think the government kicked heavy industry out of the country (to China no less) because they truly believed in short sighted mid 20th century environmentalism? Of course not. They did it because Wall Street stood to make an obscene amount of money piecemeal selling that part of the economy overseas. Environmentalism just provided the excuse and pool of useful idiots from which to build political support.
It's kind of disappointing how few efforts of the level of AlphaFold are coming from the AI research labs. I used to believe that it was an indicator of AI still being mostly useless for research, but that is obviously not the case given all of the recent contributions to Math.
I'd have expected them to be going much harder at collaborating with universities and national labs to get access to their many petabytes of curated scientific data.
It's still a skill for now. Prompt engineering specifically is perhaps not as much of a skill anymore, but I think there's a level of skill involved in understanding and/or anticipating when a model is going off the rails.
Some people struggle to effectively use AI because they either have to spend a lot of time reading and thinking about the response or they have a hard time noticing when the model is subtly going off the rails. Others use it to good effect because they can anticipate which tasks would be better handled manually, or are good at catching that the way the model is describing something subtly indicates a misunderstanding.
People call grok that because deviants were abusing grok's ability to edit images and post them publicly on X to strip people - including children - of their clothes, from their public photos. Then when there was backlash, Elon laughed it off. It took half the world opening investigations against X for violations of existing regulations for action to be taken.
The leniency that internet companies get in terms of dealing with illegal content comes with the expectation that they're making reasonable efforts to control the distribution of said content. X, and Grok, were actively supporting the production and distribution of the content in public.
It is very different from someone creating such images in a private account, and definitely very different from someone using a local model to do it.
My entire post was how it is unreasonable to have a dual standard, and my point was it's really irrelevant if it's a model you download and use locally, or if it's a model hosted remotely, or hosted and created remotely. You're not really providing any sensible reason where the line is, except "public company", which is, again, the entire point I'm making.
The only realistic, non-double standard is that the creator of the model should be 100% responsible. What on earth does it have to do with who's hosting it?
And by this metric, aren't all the uncensored models on huggingface, child pornographers? And if so, why not? Provide tangible, real, sensible reasons please, and after all, isn't hugging face a company?
You know, people are all over the place on this. I see people complaining about guardrails, then in the next breath complaining there aren't enough. Complaining that open models are the thing, but then creating double standards.
So once again, what is your actual reason why it's different?
It isn't about "public company", the line is "posted to a public social media platform, with the owner indicating amusement".
If someone uses something one of OpenAI's hosted models to generate illegal content, it isn't automatically spread to anyone that clicks on the post the original image is from. On top of that, OpenAI does not explicitly endorse the use of their models for such activities the way Musk did.
If someone uses a local model to generate illegal content, same thing applies.
I have a hard time believing that you're arguing in good faith. The differences are so blatantly obvious and they don't even require any discussion about guardrails. Grok was being used to edit images and post them on public social media. The technology to edit images has existed for decades now. So here's a direct analogy via Photoshop:
1. In the case of a local model: Someone using a local image editor to edit photos into illicit material and manually distributing them.
2. In the case of a hosted model: Someone using a cloud image editor to edit photos into illicit material and manually distributing them.
3. In the case of Grok: Someone using a cloud image editor to edit photos into illicit material, the cloud image editor automatically distributing it to the public, and the owner of the editor endorsing the material as an exemplar of what their editor is capable of.
In 1 and 2, the editor (ie the AI) was just a neutral tool. In 3 the issue is that the company explicitly endorsed and distributed the material, no longer neutral. This is why it's reputation is of being a CP machine.
The leniency that internet companies get in terms of dealing with illegal content comes with the expectation that they're making reasonable efforts to control the distribution of said content.
Your words. Grok was an example, the owner is irrelevant in the discussion.
Really showing that you aren't engaging in good faith and your repeated incorrect interpretations of my words are out of a desire to cover for Musk.
To spell it out for you further, the owner's actions matter. It doesn't matter that it was specifically Elon doing it. The owner promoting the ability to produce and post illicit content is the opposite of making reasonable efforts to control it!
At this point I just have to assume that you're trolling. Idk how much more explicitly I have to say that it doesn't matter who the owner is, it matters what the owner did.
And yet you incessantly go on about Musk, and talk about me as if I'm trying to protect him. You tried to claim that in the last post, right at the top, so yes, you're talking about Musk.
And I've now been immensely clear, even though it should have been clear from my very first post, or at least the second, that in this discussion it's not even remotely relevant what the owner of any company did. Why would it matter? We're talking about local vs corporate compute for models here, and about who creates what baked-in restrictions in models. Not about what some random owner did.
This discussion is about open versus closed models. About them being legislated. About them being controlled. Why are you blathering on about Musk? Why do you keep going on about the actions of any CEO of any company, when I was discussing corporate vs individual liability in these matters? Even now, you seem to think it's relevant what the owner of companies do here. It's not. At all. Why on earth would it be?
I feel as if the largest discussion the world is having today is, "sweep it under the rug". You seem to be trying to say, and don't claim otherwise, as you went on and on about how corporate liability is different than with private models!, that what a person does on their personal computer should have different liability implications than on a corporate leased computer.
That if content creation cannot be seen, well then all is well and OK. Just sweep it under the rug, and all will be fine!
At one point you tried to shift this into the realm of "no it's only once it's posted", and that's moving closer to a correct target of personal liability and responsibility. But these issues are now so "in the faces" of the "normal joe", that the entire internet is about to change.
And the reason for that is, personal liability is coming to the internet.
All the age and identity verification blather is about, as I've been saying here and there for years, personal liability and responsibility. It is entirely ridiculous and irresponsible to place liability for user content at the feet of the makers of tools. Instead, if you post CSAM, or if you make a fake video to slander someone(eg, to screw with elections), or if you incite violence, all of it, one needs to be held personally liable and responsible for that.
That's where the internet is going. Anonymity will continue, of course, when you post to X, to Youtube, to Facebook, your identity will not be immediately disclosed. But they'll have it, as a requirement for posting anything, and a simple warrant will disclose it. Otherwise they'll face criminal liability.
I believe 230 carve outs will change to ensure that this degree of ID verification is a prerequisite for any form of liability exclusion. And really, seeing today's landscape? This is the most free method to continue on the path we're on.
It's the most free, as the concept of forcing companies to be the filter is horrible in comparison. If companies are the filter, guess what happens? They are politically bent, or they alternatively are legal liability concerned to the point that they over-filter. And if companies become "the filter", then that's where true freedom of expression disappears.
Yet if companies are not legally liable, then the weight of liability must lay somewhere. And it should lay on the end user's shoulder. That's how you retain the ability to express, to post political satire, it's where you get the ability to protest.
Imagine a world where the roads were all owned by private companies, and if you wanted to protest politically, you were not granted the right to use those roads?! Not even out of malice, but due to corporate liability fears.
That's what corporate liability looks like. And I can assure you, 100%, liability is here.
So your thoughts that companies should be treated "special" only means one thing.
It means the illusion of freedom, whilst your ability to exercise that freedom vanishes.
Depending on where you slice the model up, it can be not a whole lot of data. For instance each transformer block outputs a single vector in an embedding space.
I can see that being cheaper to bitbang with PIO than to actually compute.
There's certainly some latency stack up, but throughput should be remarkably good.
reply