DrScientist

In person, pure paper voting, in local community halls etc, run by volunteers is the best voting system.

Incredibly resistant to influencing at scale, resistant to voter coercion, transparent, and invokes a sense of occasion and community spirit.

The problem with electronic systems is the public can't see the piles of votes, can't get involved in the count, and ultimately it's the perfect system for being able to tip the scales subtly.

Even vote counting machines for paper ballots are a mistake in my view.

Is it me ( and my eyesight ) or is the text a bit fuzzy in their example videos?

Are you treating the LLM output as machine code or are you looking at it?

And if you are looking at it - do you insist on certain patterns or structure to allow yourself to more easily navigate it?

> and decided I'd rather just use vanilla css/js for new projects.

So is the rise of AI coding, the end of frameworks?

Frameworks were created to reduce boiler plate ( at the expense of new abstractions which occasionally leak ) - boiler generation is less of an issue for AI, and the underlying platform is very well specified?

I guess the only problem with vanilla is if the sheer volume of text overwhelms the context?

I said if they were serious about 'safety' then given by their own graphs that Sol is less aligned than Astra, they should have pulled it.

As they haven't, the obvious conclusion is that safety isn't the main driver here.

I'm just following the logic - you are the one making assertions about motives that contradict the evidence.

Now I have some sympathy for the idea that these companies need to be responsible, but in the end - because these are universal machines - the regulation is going to be needed at the machine real world interface, not at the model itself.

This is, by the way, why OpenAI is responsible for the hacking that it's AI driven agents ( not models per se ), did.

It's not a problem to have an LLM that thinks it can drive, the problem is putting one it an actual car and putting it on the road.

So you are telling me Sol can't be used for hacking?

Please...

Also think about it - they are 'announcing' a delay.

They didn't need to do that - they can just release when ready - nobody is expecting them to release a new model on a certain date.

The whole thing is confected.

Things can be regulated in sensible countries where the things are pretty much single use - like guns.

However, nobody bans cars because somebody has driven one into a crowd, or hammers because of a particularly grusome murder.

These models, just like computers, are universal tools by their nature - they don't fall into the gun category - they fall into the hammer and car category.

From their own Astra press release.

>Compared to GPT‑5.6 Sol, which without production safeguards went beyond the authorized target 48% of the time, GPT‑6 Astra did this in 0% of cases.

Is Sol still available?

I assume the production safeguards ( filtering prompts and responses ) is fairly transportable between models.

The other point is if their new model isn't safe, doesn't that mean all their previous models aren't safe as well?

ie if they were truly being 'responsible' they would pull the lot....

That's what makes people think it's not genuine.

Models like this are in effect dual use - just like computers and computer programmes in general. Ethercap is a very useful debugging tool - and can also be used for exploits.

We don't ban Ethercap, we police the use - just like with all dual use stuff.

If OpenAI is directing the software ( how ever badly ) that's hacking - OpenAI is responsible, it's not the fault of the software.

Note if somebody else is using OpenAI models to generate code to hack - that's not OpenAI's responsibility it's the person using and running the generated code.

In it's heyday Sun shone very brightly indeed and was key to stopping a dismal Microsoft hegemony in the late 1990s.

In the end commoditisation eat it's lunch, just as with SGI, and they didn't manage to adapt in time.

A lesson for AI companies perhaps....

Did the police have the appropriate license to play those disney songs in a public space?

Indeed.

The tricky part is that social peer pressure requires a lack of anonymity - not just in terms of detecting whether people are recording without consent at the time - but also the posting of content online.

Naming and shaming requires names.

It's European countries that are taking this strategy, any Russia sabotage is aimed at deterring this widening of the conflict - not to encourage it.

Originally Europe didn't want to supply tanks, then it was only short range missiles, then it was longer range missiles as long as they weren't used to attack targets inside Russia - now it's actively supporting doing military strikes deep into Russia, and openly talking about European boots on the ground in Ukraine ( not that they likely didn't have deniable special forces there from the beginning ).

Strategically it's not in Russia's interests at all to have a wider escalation right now - they have their hands full. Any Russia sabotage is a direct response to missiles and drones using Western technology blowing up warehouses and oil installations deep inside Russia - widening the war is not in their interest or their aim here - they are trying to deter that.

There is the wider question of whether Russia's long term goal is to take back the old soviet states bit by bit in a way that doesn't trigger a full response - but I'd argue there is already a clear red line there - and that's the Nato border.

Not sure how MH17 shooting is relevant here.

While the consensus is that it's likely a Russia controlled/allied battery shot it down - I'm not sure anybody thinks it was anything other than an accident, not an intentional act of sabotage.

The ideal scenario is everyone suspects it's you but can't prove it.

You have both the desired threat, but no concrete biases for an official response.

Also an uncertainty limits action - it might be a random event, it could even be a false flag.

One of the problem's here is narratives are being spun by multiple actors and it depends on who you believe.

Frankly if I was China, I'd be conflicted about the mid-terms - Trump is currently leading the US to disastrous wars - kinetic and trade, pissing off his closest allies, under-mining the dollar, and making China look a sane and stable partner.

On the other hand the world might be better off if he was more constrained post mid-terms.

Frankly, in terms of countries wanting Trump to take a beating in the mid-terms, I'd put countries like Canada and most of Europe way ahead of China there.

And in terms of China's energy strategy - China is the worlds biggest net importer of crude therefore China's strategy is going to be focussed on managing that first and foremost.

I think it's the US who hoped that the oil crisis would impact China much more - this is the US making, not China.

For quite a long time, graphics card driver quality and performance has been much better on the PC driven by the larger PC gaming market.

Very hard for Apple, or Linux to do much about that - especially when Nvidia considered it's drivers key IP.

Couple that with a relatively small market - and it's hard to justify supporting multiple platforms.

Eh? They didn't/aren't using their stockpiles to anywhere near the extent the US has! There reserves are much bigger - they have about a years supply - yet they used less in absolute terms than the US!

What they did was aggressively reduce demand by accelerating the switch to EV's and firing up a few more coal powerstations.

In terms of influencing the midterms - I suspect China isn't anywhere near the front of the queue in hoping Trump get's a beating given the chaos he is causing.

Complaining about foreign interference is ironically straight from the Hillary playbook.

Using that lens then you could view the Iran war is simply a failed ( so far ) attempt to grab Iranian oil revenues, perhaps emboldened by the relatively quick success of the Venezuela adventure.

The truth is probably much more multi-headed - though hard to see how the US would have gone to war without some assessment of their own energy situation.

Blaming the current situation on the actions of China is beyond bizarre.

The original posters premise was the way to make government dragnet surveillance banned was to do a public one. In general it's a good principle however I'm not sure in this case.

I can see how a public flock, which is just a queryable database, would be ok under the current rules. But would it make the government want to regulate? They could just continue to go after data use where legal asymmetry already exists.

You could frame almost any use as some sort of stalking/harassment.

Even in this case if they shut it down, government carve outs would apply.

Sure.

That's a problem with the whole AI space - it's eating itself at a incredibly rate.

Ultimately if they add value and behave well they might carve out a market - but anything in the software space right now is a gamble.

As far as I can see there is no clear pricing once you go beyond the free tier - it's contact sales.

This is actually a classic enterprise software approach - as there are no fixed per unit production costs to cover, the pricing is very elastic.

ie each license is probably a negotiation - if you see your app for a one off fee then they want a one-off cut - if you sell a subscription they may want a cut of that instead.

Obviously the risk is if you bake it in and your app becomes successful they may want a larger cut than you can afford, but if they don't negotiate in good faith they won't last very long.

Your Harry Potter case is a good illustration.

The world Harry Potter universe clearly borrows heavily on past literature - whether that be tales of goblins and wizards or tales of boarding school.

So much so ( and the fact that she is very rich ) she has been sued multiple times for plagiarism - all of them failed.

Is it really beyond the wit of man to come up with something similarly new?

In terms of fan fiction - is it ok to do one, and promote it on the internet, that has incest themes or paedophilia? These potentially reflect back on the original - and surely the creator should have control here.

There is always the option of coming to an agreement with the original author - that's how films get made - but frankly there is an almost infinite creative space - I don't see why people's lack of creativity needs to be promoted above the rights of an existing creator.`

Note in the example you quote - JK Rowling has allowed the non-commercial publishing of that and other fan fiction - just blocked independent commercial exploitation. So in the case you made the existence of copyright didn't stop Harry Potter and the Methods of Rationality.

I'm trying to distinguish the collection of information from the use. Collection is legal - however there is no point to collecting it if you can't actually use it for anything.

For example:

https://www.pbs.org/newshour/nation/federal-prosecutors-char...

or https://www.justice.gov/usao-cdca/pr/santa-monica-man-pleads...

Another area that's also existed for quite a while is the crowd-sourcing of the positions of mobile speed traps.

Though bizarrely in the UK that's legal if you do it via Waze, not if you do it via whatsapp.

It's legal to film publically - what's not necessarily legal is what you then do with that.

There are already precedents - for example it's legal to film ICE agents in public spaces, it's not necessarily legal to use information to dox if that can be construed as harassment.

Even if you crowd sourced the identification of somebody involved in a potentially criminal act, you can only share that with the police and not the wider public.

The law is asymmetric here because the police have powers reserved for themselves to stop vigilantism.

Perhaps an interesting angle here would be if you can make the case that companies like flock are involved in harressment themselves...

Good point.

By the same argument, wouldn't all the things you don't like about copyright - like restrictions with what you can do with the thing you have bought - just be put directly into the purchase agreement.

Isn't the way to focus on creating consumer protection rights, rather than just leaving it to people with the best legal department and deepest pockets to control the rules?