HarHarVeryFunny

I wonder how the "Chinese AI is all just distillation" folk are going to cope now that the Chinese are all aboard the post-train via agents in custom training environments wagon?

Here's Xiaomi's discussion of this, plus their open-sourcing of 7000+ RL training environments.

https://www.alphaxiv.org/abs/2609.mimo-scaling-reinforcement...

https://mimo.mi.com/docs/en-US/news/latest/v2-6

Who needs a few of someone else's "vacation postcards" of their post-training experience when your agents can go on vacation themselves!

It seems to be a mix of LLM + RAG, with the RAG search results preferring what's fresh vs what has at least been endorsed by pagerank.

A large portion of the "New LLM release" story comments here on HN also appear to be astroturfing ... would be interesting to see a correlation of comment sentiment to IP-derived geographic location!

Google's AI mode is extremely hit or miss. At it's best it's a more flexible version of their old pagerank "I'm feeling lucky" button, where the LLM does return an appropriate response ...

However AI mode seems to be largely RAG, so it's supposedly authoritative answer is often just some online forum comment it picked up, such as a couple days ago when it fed me back something I had just posted myself 30 min earlier!

It does at least dump you into Gemini chat, so when the initial answer is obviously wrong you can try to drill down to get something better, but presumably many people are taking the initial answer as some sort of "AI truth" when it is far from it.

The EDSAC's "boot ROM" consisted of a bank of large electro-meachanical "uniselector" rotary switches, hardwired to contain the 31 instructions ("initial orders") of the tape loader.

Subsequent computers could have followed this "electro-mechanical ROM" approach, but I suppose it was rather bulky and inconvenient to change.

You can see what the uniselectors looked like here at 8:50

https://www.youtube.com/watch?v=Yc945sNB0uA

Of course programmable ROM chips would have to wait, so in the meantime the alternatives were either the toggle-it-in approach (still also used by the 1974 MITS Altair 8800), or other more compact types of handmade/hardcoded physical ROM such as magnetic core or diode matrices.

Saying "like any other stored program computer" makes it sound like there were a bunch of them, but as I understand it there was really only ONE up and running at that time in 1948 - IBM's SSEC. The stored program EDVAC, whose architecture Von Neumann had formalized, was still not yet running, and Von Neumann personally advised on adapting ENIAC to this mode of operation.

As I understand it, even after adaptation to "stored program" mode, the program was still configured rather than loaded and stored, but now configuration was done using the function table dials (essentially a dedicated dial for each program step) rather than patch cords, and was a lot more convenient. Punched cards could be used for reading/writing data, but not for programs.

It seems there was no BIOS because I/O was so primitive. Per configuration a punched card would be read into the "constant transmitter" unit, then one of the read numbers could be copied into one of the 20 accumulators. Or you could do the same in reverse to output data from the accumulators.

It's perfectly legitimate to investigate any theory in the scientific manner of falsifiable experimentation, but I'm not aware that can be said of anyone looking for a quantum explanation of consciousness (which is anyways a rather odd quest given that there doesn't seem to be anything needing explanation).

Roger Penrose certainly appears to be a quack when it comes to quantum consciousness and/or brain operation - it seems that he wants to believe it, so waves his hands vaguely in the direction of microtubules as a place it could be hiding, despite zero experimental evidence, or any falsifiable theory. It's not the fact that this flies in the face of everything we know about quantum effects at classical scale and room temperature that makes this quackery - it's the fact that this is just wishful thinking, not science - not the application of the scientific method of falsifiable theory followed by experimentation.

The article acknowledges that quantum behavior like tunneling necessarily happens everywhere, if you look at the relevant scale, and also seems very clear about what it is actually talking about - whether classical-scale quantum effects can be seen in biology.

You seem to talking at cross-purposes to the article - not arguing against the research it is discussing of classical explanations of quantum-like behavior, but wanting to point out that biology is not immune to quantum effects at the scale where that is not surprising.

Whether Jev is something architecturally different from an LLM (bidirectional vs auto-regressive? different type of parallel readout head?) remains to be seen, or guessed, but LLMs are certainly fungible and are competing on price - they leapfrog each other from release to release, but overall they are all progressing in unison.

When it comes to the high volume market of business automation, it seems that ultra-low cost, rather than expensive frontier intelligence, is exactly what you want, and low latency is also nice to have for customer-facing applications like customer service chatbots.

I didn't even realize what I was looking at to begin with - I thought it was modern sculpture looking very interesting and organic, although the "installation" locations seemed almost impossible! :)

It seems people are just guessing at the architecture behind Jev. Obviously the functionality itself is easy to replicate, but why Jev seems to be making such a splash (beyond the doh! factor of it's huge applicability) is the ultra-low cost and speed, which may be due to architecture.

The Laya model compared in TFA shows one way Jev may be getting it's speed and low cost - by using a BERT-like bidirectional model rather than an auto-regressive one (LLM).

Their self-interest is in keeping Putin at arms length and avoiding WW III. Putin is already threatening Lithuania, sending drones into Poland. If Putin wasn't prevented from taking Ukraine, do you really think it would end there?

It's hard to see Putin ever choosing (under pressure) to end the war unless he can at least claim some sort of victory, such as territory gained, although never say never. The Soviet Union eventually just accepted their loss in Afghanistan and withdrew, although it took new leadership to do that. For that to be politically possible probably requires the war to become felt even more at home in Russia, and become more unpopular, which seems to be slowly happening already, even before more advanced missile systems become available to Ukraine.

They have no NATO obligation towards Ukraine, but I'd say they do have a moral one, and they also have their own self-interest.

Sure you can accelerate the regulatory process, but with drugs it's not always going to work out well. Look at something like Thalidomide - sure would've been better if approval had taken longer (and been denied).

Sure - I agree with Anthropic's position here, but at the same time it makes sense that the military doesn't want to use them, and there are certainly situations where the government needs to be given the powers to do what they are elected to do, but hopefully limited to extraordinary justified situations.

What if the private company had a vaccine to prevent a pandemic?

The job of a government is to look after the interests of the people who elected it... Is there no point that you think the rights/needs of the many should outweigh the ideological rights of the individual?

The parent was talking about the disunity of Europe in standing up to Putin.

What would come next after Ukraine may be the Baltic states or Poland.

They don't need to - there are alternatives, and do you really want to do business with a supplier that is hostile to your needs unless you have to?

Maybe, but only in a very idealistic, unrealistic, sense.

Could you imagine a hypothetical where we're at war and a private company has a potentially war-ending tech (nuclear weapon say) that they refuse to sell to the government?

> The subject at hand is an attempt to use national security regulatory power to coerce Anthropic into selling a product it doesn't want to sell.

It would be naive to think that a private company could deny the government access to any tech if they want it for military or surveillance etc usage. In this case the government isn't forcing Anthropic to do it, they have alternatives and are perhaps making an example of Anthropic for refusing.

You could even make a case that the "supply chain risk" designation DOES make sense. If Anthropic are opposed to certain uses of their software, then doesn't that potentially impact companies that are part of the supply chain as well?

Evidentially not - Anthropic want to do business with the US military, but on their own terms, telling the military what they can and cannot use it for.

Supply chain risk seems a massive over-reaction, but understandable that the government doesn't want to commit to Anthropic controlled software for any critical military application if they can't 100% rely on Anthropic support in being able to use it as they see fit.

Because NATO article 5 demands it.

European support started out slow, but has ramped up as the US has backed away.

Germany will spend 20% of it's federal budget on military in 2027, planning to increase that to 30% by 2030. "We cannot defend ourselves against Putin with a balanced budget".

> But now Trump is inviting Putin to G20 and I'm sure sure all those EU politician would be all smily and act like no war is going on and gladly talk with Putin

Putin is deathly afraid of being assassinated and is avoiding air travel in favor of trains, and has identical offices to film in at multiple locations to hide his location. There is no chance he is getting on a plane to Florida.

> If you think about it the war could have ended years ago and all the Ukranian/Russian lives lost could have been avoided if EU leaders were not so arrogant.

How so? By Europe rolling over and not supporting Ukraine?

Try googling for information and use whichever news sources you prefer - just don't expect to find Fox News talking about it.

If it comes down to European vs Russian manufacturing capacity for missiles and hi-tech weaponry, etc, I'd put my money on Europe. Europe also has the advantage of seeing 5 years of Russia throwing their best at Ukraine, having their air defenses repeatedly breached by Ukraine, even deep into Russia using slow low-tech drones - they know many of Russia's weaknesses.

Ukraine is only Putin's first step, and Europe is well aware of that, so look at all the advanced weapons systems Europe is now giving to Ukraine (Storm Shadow, SCALP, SAMP/T etc), as well as the statements they are making. They are well aware this has made them a target for Putin, with UK missiles having been used against Black Sea targets etc, but are doing it all the same.

Europe is not tolerating Russian drone incursions, and has been shooting them down, have been on air combat footing patrolling the finish border and Kalinograd region (operating out of Lithuania), boosting defense of the Suwałki Gap, etc.

YouTube is perhaps one of the best sources of information, since US MSM is not interested.

Europe has been stepping up a lot recently, given Trump's reluctance to push back against Putin - looks at UK/France supply of missile and anti-missile systems and IP to Ukraine, stepped up air defense based out of Lithuania (anticipating Kaliningrad as the first front of the war), etc.

The only way to stop Putin is with military force.

Europe is of course well aware that war with Russia seems increasingly inevitable, and have been preparing for it.