deepnotderp

Shelter and gasoline are self inflicted injuries

Looser zoning, land value tax and investments in energy would solve them

That's because CPU performance cares less about transistor density and more about transistor performance, and 14nm drive strength was excellent

I’m not a fan of Extropic, but I’m seeing a lot of misconceptions here.

They’re not building “a better rng”- they’re building a way to bake probabilistic models into hardware and then run inference on them using random fluctuations. Theoretically this means much faster inference for things like PGMs.

See here for similar things: https://arxiv.org/abs/2108.09836

There’s a company called Normal Computing that did something similar: https://blog.normalcomputing.ai/posts/2023-11-09-thermodynam...

More rotors aren’t more efficient: the more rotor area you have, the more efficient.

Whether you achieve that with 4 small rotors or 1 big rotor is immaterial.

The reason eVTOLs use many small rotors is because an eVTOLs only operate in vertical mode for a short period of takeoff and landing. They aim to be normal fixed wing aircraft for most of the flight!

A helicopter like big rotor would be inefficient- adding weight and drag during forward flight. Helicopters on the other hand, spend the entire flight in vertical mode, for which a big rotor is most efficient.

this stems from newtons second law, but is neglecting second order effects like interference

> Yeah but then you get mugged on the street daily if you flash your phone.

I agree, that’s another good reason to avoid sf

> “ But then somewhere around 10nm everyone just gave up and started handing out random numbers.”

Hahaha

> GPUs Groq will still come out on top because you can never reduce latency by adding more parallel compute :)

You literally can, in fact that’s the entire reason to use multiple chips.

See eg the TPU group’s paper: https://arxiv.org/abs/2211.05102

That article is the purest example of how college has become about signaling and not substance.

someone does something commendable in STEM

MIT: “We didn’t let them in, which means it’s bad since we are the arbiters of what is good and bad”

This demo has more than 500 chips btw, it’s not exactly an apples to apples comparison with 1 GPU…

There are three routes to bypass this issue:

1. Move to supercritical Co2 turbines (lower mass —> lower cost), need RnD on corrosion resistant alloys

2. Move to thermoelectric or photovoltaic generators

3. Use aneutronic fusion

Directly use the charged particles

> Do you think that speech is not an act? That speech does not have any consequences in the world, and so should be free from all restriction?

YES

(Other than incitement of violence and libel)

Single transistor level frequency is different than chip level frequency.

Each clock stage has many layers of transistors

> SSRIs have been subject to extensive testing and approval processes.

SSRIs barely beat placebo and have a significant side effect profile

I always find it interesting that whenever an article about psychedelics gets posted, the comments are always “We don’t know yet if this is positive, we need much more evidence and science!”

Yet the standard of rigor for alcohol, weed, SSRIs and tiktok is somehow not as stringent.

Do we hold things like alcohol or tiktok to the same level of rigor?

> when we need science.

Do SSRIs beat placebo?

Do… do you know how psychedelics work?

Agreed, and we see in this in particular with the fervor to try and create “non hallucinogenic psychedelics “

Associated != causal

Most people who are vegan are more likely to be health conscious

> Can’t get so fat and unhealthy. And far healthier since most vegans and vegetarians live long and healthy lives compared to other groups.

My healthy breakfast of donuts and a soda ;)

Jokes aside, I don’t think the health point stands, but the animal health perspective probably does.

Much in the same way that the “Wright brothers flight has nothing to do with transportation”

> This design is called the Teller-Ulam design, and it is extremely likely that it was discovered only once (by Ulam and Teller), all other thermonuclear bombs are just the result of successful spying and secret sharing.

Hans Bethe believed that the Soviets and the British got it by analyzing the fallout from fusion bomb tests. Apparently the high number of heavy elements suggested high rates of compression

Are modern petawatt class lasers really 20% efficient?

>“ It may be wonderful if we could, but at the moment, every dollar spent on fusion is a dollar not spent on building more solar panels and wind turbines.”

Every dollar spent on cars and transistors is a dollar not spent buying more horses and vacuum tubes

Something something transformers

Doesn't sub-quadratic attention fundamentally imply forgetting things?

Nvidia’s response is the shady one, in their response they’re using a different batch size

Most HPPD is just floaters or tracers, and SSRIs can cause them too.

The magnitude of HPPD and the frequency/prevalence need to be studied, but I think seeing some extra floaters in exchange for having a will to live is a trade most would take.

It’s not for a lack of trying on Samsung’s and intel’s part that tsmc is dominant ;)

NIL has

a) Too high a defect density

b) Too poor alignment between layers

to directly replace EUV

4-aco-Dmt and shrooms are both pro drugs of psilocin so they’re functionally the same molecule

DMT can be life changing to many people, me being one of them, but it’s also extremely powerful, scary, and not always a positively life changing experience

EDIT: I misinterpreted what you meant by the 4-aco-Dmt comment. I actually agree that it’s better than actual mushrooms, especially for beginners

2C-B is another less overwhelming option for beginners