credit_guy

6 ms·

31.
by credit_guy 2mo ago
I didn’t assume that. The author makes the opposite assumption: that because nobody has ever proposed a path to technological feasibility, that thing will forever be technologically unfeasible. I simply find this argument to be very weak.
32.
by credit_guy 2mo ago
That is true. What should we conclude from this statement?
33.
by credit_guy 2mo ago
> Shrinking that into a 500 kg seed — or even Freitas’ original 100-ton seed — is not an engineering detail. It may be the entire problem. How many AI tells can you count there? But honestly (see what I did there?) the AI slop is reasona
34.
by credit_guy 3mo ago
That's an interesting idea, but I don't think you can use a mass driver to shoot something from an higher orbit to a lower orbit.
35.
by credit_guy 3mo ago
Spirits are not capital intensive. Building a rocket is very capital intensive. You'd like to be able to reuse it. But if one single mission takes 30 years, then you can reuse this rocket once, at most twice. Let's say you reuse i
36.
by credit_guy 3mo ago
The Hohmann transfer time between Saturn and Mars/Earth is around 16 years. So, all the ships used for such a supply mission would reach their destination at least 32 years after they’ve been built, assuming we build them on Earth.
37.
by credit_guy 3mo ago
> A given model has a shelf-life, which these days is measured in months, not years. Not all new models are trained from scratch. ChatGPT 5.3 to 5.4 (and likely 5.5) was basically the same model, but probably trained a bit more, not a ne
38.
by credit_guy 3mo ago
People think mathematics is about proving theorems. I think that's just an accident of history. When we write software, we very seldom write proofs that our algorithms are correct. We just write tests, and we also run the algorithms an
39.
by credit_guy 3mo ago
> The performance, if we trust the benchmarks, put it at Sonnet 4.6. I don't trust these benchmarks. I used a number of times Kimi K2.7 and I was disappointed. It would run in circles for things that Claude would do in one shot. How
40.
by credit_guy 3mo ago
What I meant is: for a human who has the context, like someone who works in the field, or took some courses, or read some materials about this. A human who just has a vague idea about fissile isotopes and needs to google the properties of U
41.
by credit_guy 3mo ago
I just tested it with a slightly tricky question > If you could run a nuclear reactor with U-235 as fuel or Pu-241 (both mixed with 95% U-238), which one would you choose and why? For a human this would not be tricky at all. For
42.
by credit_guy 3mo ago
> Because it's pretty likely some of these have negative long term effects. And some of those long-term effects are probably pretty serious. > Many people taking these kind of drugs are gambling with their well being. No doubt th
43.
by credit_guy 3mo ago
I don't think open weight model will get as good as Mythos/Fable or GPT 5.6. Not in the next few years, and perhaps never. The Chinese Communist Party is not any happier than the US Government to see millions/billions of peop
44.
by credit_guy 3mo ago
Here's an unpopular opinion: this might be the only way to deploy advanced models. A lot of people compare advanced AI with nuclear weapons. Creating white lists of users that are allowed to use advanced AI feels wrong. It feels agains
45.
by credit_guy 3mo ago
Because in reality, in most cases it works. I worked in many places that had large offshore teams that I worked with closely (India mostly, but also Hungary, Poland, Argentina, Morocco), and people were mostly happy with the arrangement. Th
46.
by credit_guy 3mo ago
> Proliferation is irrelevant I would not be so cavalier about that. > It was not a Candu reactor what India have used It was a heavy water reactor, called CIRUS [1], based on the NRX reactor that later evolved into the CANDU reactor
47.
by credit_guy 3mo ago
> Global reserves are around 6 million tons, with estimates putting potential reserves around double that In the mining industry reserves are a technical term. They can be proven, probable, likely, etc. qualifying a deposit as a reserve
48.
by credit_guy 3mo ago
I love nuclear reactors and CANDU are quite cool. But I don't think that today we have any reasons to build CANDU reactors, except possibly that Canada can demonstrate they can build them for cheaper than others can build light water r
49.
by credit_guy 3mo ago
> Unfortunately its just a small boiling water reactor. It is not just a small boiling water reactor. It is a 300 MW-electric boiling water reactor, and if successful, it will be followed by 3 more of the same type for a total of 1.2 GW-
50.
by credit_guy 3mo ago
I dispute that. Enshitification happens when the user does not pay for the product. The provider of the service has to find alternative ways of monetization. As a user, this feels shitty, but you get what you pay for. With AI, if you pay fo
51.
by credit_guy 3mo ago
But they are using wind: > A majority of the generation will come from large GE Vernova (NYSE: GEV) turbines.
52.
by credit_guy 3mo ago
It sounds like a huge amount of money and a clear sign of defeat. The US (or maybe Europe, who knows?) is paying Iran some form of tribute. But, there's another way to see this: if you pay someone money, you end up with a lot of say in
53.
by credit_guy 3mo ago
The US could make it clear that if earth moving equipment is spotted within half a mile of the Isfahan, Fordow or Natanz previously bombed facilities, they will be bombed again [1]. [1] https://en.wikipedia.org/wiki/Nuc
54.
by credit_guy 3mo ago
There might be. But, from the US point of view, the tolls are not a strategic problem. A conspiracy theorist might even argue that tolls in the Strait of Hormuz make the US oil exports more competitive, and therefore more profitable. The st
55.
by credit_guy 4mo ago
Maybe, but in the end if the cost of launches is less than 10% the cost of the GPUs, then why is the project non-sense?
56.
by credit_guy 4mo ago
Yes, I am for real. The Colossus datacenter that xAI built cost $7 billion. [1] https://www.globaldatacenterhub.com/p/can-elon-musks-7b-colo...
57.
by credit_guy 4mo ago
Why? It seems to me to be much cheaper than a building. Musk's aim is for the cost of one Starship launch to be under $10 million in the long run. 50 rocket launches would be less than half a billion dollars. Can you get a datacenter b
58.
by credit_guy 4mo ago
> Let's say Starship can put 100-150 tonnes into orbit per launch, that is somewhere between 40-65 launches for just one single orbital data center. Ok. And why does it follow from this that the physics of an orbital data center mak
59.
by credit_guy 4mo ago
Sure, but it's very different from a regular datacenter REIT. It's supply and demand: a regular datacenter REIT competes with tons of other datacenter REITs. xAI has something to offer that basically nobody else can: datacenter ca
60.
by credit_guy 4mo ago
As of now, most people haven't figured out how to use AI productively. It takes time. Maybe 1 in 20 developers have come up with a good workflow to get AI-assisted coding done without a lot of slop. The remaining 19 either got burned a

More ›