Northern Mage 🇨🇦 🇬🇱 🇩🇰 🌍

Oh ye of little faith.

Trump can ALWAYS find ways to make it even worse for America.

And he always does.

Or have their browser histories made public

It seems that Trump and his cabinet are dedicated to speeding up the arrival of the post American world

of the Ukrainian people to not be conquered and the surge of support from other countries.

I fear that the US may be repeating the Russian blunder of overestimating their prowess and underestimating their enemy. 3/3

stand a chance against the overwhelming military might of Russia. Their assessment was based on the number of tanks, planes and weapons each side possessed.

The metric did not count the condition of the assets (falling apart), the morale and motivation of the troops, the sheer determination 2/3

The other factor to consider is that the progress of a war cannot be determined by a single metric. The focus on airpower and missiles might blind the US/Israel to other strategic moves being made by Iran.

It was only a few years ago that I remember all the pundits claiming that Ukraine didn't 1/3

The USA has become the obese spoiled rich kid who is the playground bully.

"Shooting. Hiding." Same difference to US CENTCOM

She's a Trump appointee; of course, it's criminal negligence. Duh.

And he has absolutely no idea that he is

It's a sacrifice that Trump and his Epstein cronies are more than happy to make.

If this clown car of idiots in the US government were capable of rational thought, I would guess that is the point. Like Gaza, destroy everything, then just annex it and take it all over to protect the people from radical extremists.

Nancy Mace is such a piece of shit.

I'm just waiting for the air strikes on New Mexico to start.

Done

So... this isn't going to happen.

Curious as to why you refer to him as "former President Trump." Suggests this is a copy-and-paste from a post made during the Biden administration. Didn't happen then, won't happen now.

*not happen

There is an old joke in the AI community that if we created an AI-based file search that acted like a human, it would make a couple of random attempts to find a file, then generate an excuse as to why it couldn't find it.

Living next to the USA makes me more grateful daily to be living in Canada.

arising from biological complexity. Accepting this as an axiom is critical to the argument that LLMs can display emergent intelligence. This materialist reductionism has been challenged across various fields.

Is AGI possible? I hope so, but LLMs are where it will happen since they are "incomplete"

observed in complex systems, however;

a. Is something emergent because it rises in an unpredictable manner or do we just call something emergent because we lack the tools necessary to predict it?

b. There seems to be an assumption by many that intelligence MUST be an emergent behaviour

4/n

Nor does it reflect all the social and discourse information we see in language use.

Language users are far more than a statistical model. And LMM is not. It is a model of word and meaning distributions in text, not how speakers use language.

3. Emergent behaviours can be

3/n

the training dataset, the more contextually precise the model becomes.

2. The categories and relationships expressed in language text are reflections of the categories and relationships used in cognition, but they do not reflect all that happens in cognition.

2/n

1. LLMs are statistical models of language that reflect the correlations between words in texts along a very large number of dimensions. The model's sheer size enables precise but not necessarily accurate text production (hallucination). The larger the number of dimensions, the more massive 1/n

"Women, Fire and Dangeous Thing: What categories of language reveal about categories of the mind." (I may have got that slightly wrong)

LLMs do not model the mind; other aspects of AI are needed for that. And LLMs do not have the capacity for growth and self-directed evolution.

LLMs cannot lead to AGI any more than improving the quality of a generated video will lead to an autonomous agent. LLMs are, at their core, statistical models of language, not language users. Why do they seem to work, as George Lakoff titled his book 1/2

This is true for all technology; the technology itself is value-neutral; it is the use to which we put it that either provides benefits or harm. AI is no exception.

Exactly right. And the other thing that people get confused about is that LLMs are NOT all of AI - there are many other areas of AI that have a track record of building useful and robust applications.

It's a necessary feature for the DOJ to identify which files to remove that they might have missed in their original purge, specifically any that could implicate someone wealthy and powerful. If the member then claims to have seen the file, the Trump DOJ can claim the member is lying.

Then it's a system that should not exist