Not futuristic: but adobe - the cheapest/oldest building material around - has this property naturally (on a 12hr cycle). Thick walls insulate from the heat during the day, and radiate that heat during the night: It's hyper efficient.

Building regulations killed it's use in America! Requiring the adding in of rebar actually makes it weaker... as well as more expensive than wood (go ahead and guess which groups lobbied for that set of regulations).

That hesitation indicates the feeling that what you are about to type matters.

Mayhapse - in the context of getting the AI to behave as you wish - such hesitations are valid. not because it is conscious: but because the context window would be polluted or corrupted... possibly mis-aligning the agent in the process.

Santa clause is not a being: modeling him as if he were can be useful, an obviously pointed example is in certain discussions about what it means to be 'real'.

My point is, if your instinct is to be kind: don't quash that because you don't consider what you are talking to as sentient. I don't yell at my rubber duck. rubber ducky is just going to rubber ducky.

I buy that.

1. To the extent that a chatbot is trained on real human interaction, we should exhibit real human interactions for best result.

2. You are either a kind person or not. A kind person behaves kindly without asking whether kindness is warranted.

The word you are looking for, when your proprioception is extended into the tool (like feeling you are the car) you use: proprioextension. coined a while ago.

Is there actually an advantage? that's toted. but no one can ever point to real data about it... and all the data suggests the exact opposite... that for most cases: cis-women out-compete trans-women.

It’s not in the article.

They list their findings but no data. They effectively are just issuing an opinion. The opinion may be more considered than the rest of ours, but it’s not data.

bad comparison - here is one better, not a perfect one...

You can't enter a car into a boating competition. The question here is: if you take basic precautions to make it the same class of boat - a modified car turned into a boat should be a valid entry - provided the engine speed roughly matches.

People worry about cars on water here, not knowing that doesn't exist by definition: any car in water has been modified from a car to be a boat. you may recognize that it was once a car - but that's vestigial shell stuff. the inter-workings are a propeller - not a wheel.

I see your argument and has some merit but isn't persuasive enough. I would posit that its a bit too loose and that it breaks down on biological people have many more complicated systems that aren't simply re-categorized similar to your car and boat comparison.

For better or worse nor is our medical science sophisticated enough to swap out the systems to be true comparables (and I don't mean to offend anyone).

> For better or worse nor is our medical science sophisticated enough to swap out the systems to be true comparables

The problem is that it isn't a hard binary. All the relevant metrics are going to fall on a spectrum, and there is a significant overlap between the male and female spectra.

The real question is: do you consider it fair if a top 1% male spectrum transitions to a top 1% female spectrum, or it only fair if that top 1% male spectrum ends up at the 50% percentile on the female spectrum?

Please do demonstrate the difference in this context.

Hormone expression. Muscle mass. Reaction time. Weight.

A YEAR of hormone therapy. Meeting a required measured threshold of testosterone.

And that's not even the controversial stuff. A man and a trans-woman are different. hell, one has (generalizing here) boobs: come on... don't be dense/obtuse! Have you tried running fast suddenly having boobs when you did not before?!?! ...one is way easier.

Oh thank you, but I’m not uninformed, and genetic testing wouldn’t have missed Castor Semenya either.

Maybe the mechanism for memory is only tangentially related to the context window.

I suspect cleverer mechanisms of context injection/pruning/updating would result in effective memory more so than my suspicion increasing the context window forever will do, regardless of what tricks we apply to distil attention over it.

There is probably a lot of low hanging fruit in this area.

I mean, or you could help build houses instead of drain electricity. One is clearly less lazy and helping humanity more.

Not paying taxes is a drain on society

It's for certain a drain on the military industrial complex, but building houses while not supporting the current regime is certainly better than draining a bunch electricity to enrich only yourself and paying money to a bunch of authoritarian wannabe's.

I am impressed with your compression of the entirety of this conversation down to two values of right/wrong. /s

You compressed all taxes down to paying for the war in Iran

The only issue. *

Reasoning is hard, reasoning about colors while wearing glasses that obfuscate the real colors... even harder... but not the core issue if your brain not wired correctly to reason.

I suspect the way out of this is to separate knowledge from reason: to train reasoning with zero knowledge and zero language... and then to train language on top of a pre-trained-for-reasoning model.

LLMs already use mixture of experts models, if you ensure the neurons are all glued together then (i think) you train language and reason simultaneously