The "OpenCooler" A silent, drop-in water cooling unit for labs, temp controlled mattress covers, pet coolers, whatever. The current market stuff is too expensive and it looked like a good learning project <https://hackaday.io/project/205182-opencooler>

I tried gamedev before, but was never able to fully commit to getting over the learning curve for Unity and Unreal. Godot has proved much easier to just jump in and make things. I really like where the project is heading. Godot + Blender = the dream

This is literally my first time hearing this. All the stuff I see from china is about lying flat, giving up because no matter how hard you work it won't make a difference? Is this a Shenzhen attitude?

There is probably something to be said about living someplace that is actually investing in itself. Seeing new development actually rise to meet the demands of the population. Seeing new transit expanded. People uplifted out of rural poverty. New technological developments. The whole bit.

The US probably felt a little like that in the immediate post war period. The enthusiasm coming out of a terrible war and a terrible depression and seeing actual changes take place in the scale of weeks before your eyes must have been something else.

But today, most cities seem to have been content with solidifying into amber over the last 50 or so years. No investments into society. The poor are still poor and objectively have worse opportunities given the buying power of the jobs available to them. Development isn't happening on a scale to actually meet the population's needs. Transit and most public good efforts are an afterthought because of no direct business profitability angle. It becomes hard to get excited about medical advances when you understand the realities of our healthcare system and that many who need these medicines or treatments won't ever get them. No enthusiasm for anything. A large population of people against anything changing. Young people and young ideas stonewalled out of positions of power in favor of people who ought to have retired by now maintaining the status quo. Technological advances seemingly solely focused on establishing new ways to rent seek, gouge, police, control thoughts, versus things that are simply beneficial to others. "no brainer" ideas facing pushback. Common sense not being valued. The optimism coming out of the civil rights era dashed away against the realities that hate towards your fellow human is a position that will carry popularity in this country. Profit above all. Control above all. Blatant corruption and cronyism by the ruling elite. Awareness that we haven't taken off the shackles of feudalism.

"lying flat" and unemployment are a thing, but nowhere near as bad as media makes it out to be. My experience is mainly in Shanghai and Hangzhou.

Organic adaption and persistence of memory I would say are the two major advancements that need to happen.

Human neural networks are dynamic, they change and rearrange, grow and sever. An LLM is fixed and relies on context, if you give it the right answer it won't "learn" that is the correct answer unless it is fed back into the system and trained over months. What if it's only the right answer for a limited period of time?

To build an intelligent machine, it must be able train itself in real time and remember.

Yes and: and forget.

Why is forgetting important? Things can either have an end time where they are no longer applicable or things we thought were true turn out to be false but it's still useful to see where we went wrong.

I imagine humans are limited by the # of synapses we have so it's useful to forget but maybe machines can move the useless stuff to deep storage until it's dug out, in the same way certain things can trigger a deep memory in humans.

Which is cheaper, among these two logically equivalent things: reducing one weight, or increasing every other weight?

Do you remove dead code? Get rid of clutter? Ever try to change a habit?

Yeah, I delete dead code but it's important to remember why I wrote it that way in the first place and why I'm deleting it now. Doomed to repeat past mistakes and all that.

Consider where that methane / natural gas is coming from. Their homepage says (with the world's most convoluted hyphen-omitting compound adjective, but basically see the last 6 words):

> March 2024: Terraform completes the end to end demo, successfully producing fossil carbon free pipeline grade natural gas from sunlight and air.

If you take carbon from the air, mix in energy from the sun to turn it into a fuel, then burn the fuel (undoing the reaction), where is the pollution?

There's more efficient ways to solve the climate problem than to install ginormous amounts of gas production, like you can run a heat pump instead of creating methane from that energy, but it's a solution that'll please even the old farts (no pun intended)

Not seeing the economics for this ever truly working out, CO2 PPM is very low despite our best efforts and the amount of energy required for separation is substantial.

The economics of taking out a loan or insurance for things you can pay out of pocket also don't work out, but there are apparently entire countries routinely buying groceries and appliances on credit

I'm not saying this is a logical thing to do (note where I wrote "There's more efficient ways to solve the climate problem"), but I've seen humans making less sensible decisions than this one so, who knows, it might actually happen...

> But they are still only going at half the pace of OpenBrain, mainly due to the compute deficit.

Right.

I honestly can't tell if this is a bot post because of just how bad I find Deepseek R1 to be. When asking it complex questions based on an app I'm working on, it always gives a flawed response that breaks the program. Where Claude is sometimes wrong, but not consistently wrong and completely missing the point of the question like Deepseek R1 100% is. Claude I can work with, Deepseek is trash. I've had no luck with it at all and don't bother trying anymore

It has a 64k context window. O1 has 128k Claude has 200k or 500K

That is exactly what TikTok does, they are more popular because their tastes algorithm is even better than Shorts for figuring out what you actually want to watch

Permits and inspections might be the the reason for humanities downfall then, at what point does war become the more efficient option?