alasdair_

3 ms·

1.
by alasdair_ 1mo ago
I heard it when I worked at Niantic right when they spun out of Google. Maybe they stole it from you :)
2.
by alasdair_ 1mo ago
Looks very useful - thank you!
3.
by alasdair_ 1mo ago
Yes, basically like alphago. I’m teaching it to play magic: the gathering. I had to start with some heuristic-based bots that played the decks very simply just to get to the point where the was some signal to learn from. I did behavioral cl
4.
by alasdair_ 1mo ago
I bought a 5090 a year an a half ago for $2000. The same card, now a year and a half older, is $4000. Then there is the RAM - I bought 96GB, wishing it was 128, and now the price on my old RAM has doubled. Stuff is crazy expensive.
5.
by alasdair_ 1mo ago
There is a lot of "AI demand" that isn't just running inference on an LLM whose weights you downloaded. I'm training a model using reinforcement learning with self-play. I can and do use vast.ai when scaling but for expe
6.
by alasdair_ 1mo ago
GCP is famous for having two ways to do everything: the undocumented way and the deprecated way. You choose.
7.
by alasdair_ 1mo ago
I still think Iain Banks was the best sf writer ever to have existed. The Player of Games is an easy read if anyone wants to try him for the first time.
8.
by alasdair_ 1mo ago
One reason is the ssd. A heavily used ssd is just slower than a new one.
9.
by alasdair_ 1mo ago
Who is going to enforce that law against you, the owner of a massive army of drones?
10.
by alasdair_ 1mo ago
Because a month later is a month of recursive self improvement at the speed of light. Once a lab catches up, the first lab will be TWO months ahead, then a year ahead, then forever ahead. After a few months of this the differences in absolu
11.
by alasdair_ 1mo ago
I’m still at the point where Fable is still very stupid and needs constant oversight and correction and questioning to keep it on task. Anything less would be close to unusable.
12.
by alasdair_ 1mo ago
You can turn off the reverting to opus with an option.
13.
by alasdair_ 1mo ago
Elon has a moral compass. The problem is that it seems to always tell him whatever he wants to do is the morally correct thing. It’s worse than no compass - his is faulty.
14.
by alasdair_ 1mo ago
Elon burned too many bridges to warrant ever supporting anything he is associated with ever again.
15.
by alasdair_ 1mo ago
Yes. The emphasis is on “more” trusted. Supreme court justice positions should be limited to around thirteen (or some other prime number) years and should have mandatory cognitive capacity tests given yearly that are videotaped and broadcas
16.
by alasdair_ 1mo ago
My wife got lost in LA and ended up in Compton. (2 decades ago). She absolutely had people tell her every fifteen minutes she was not safe there and several offered to wait with her while she ordered a taxi. Except none of the taxis ever ca
17.
by alasdair_ 1mo ago
When the president can just pick the judges he wants and then have them change laws that have existed for decades on a whim, any kind of meaningful rule of law is absent.
18.
by alasdair_ 1mo ago
The poster knows the phone numbers that were called. It doesn’t seem difficult to check who owns the numbers.
19.
by alasdair_ 2mo ago
I used to be a heavy player of MMOs, and most of the time I used a Razer Orbweaver (1 handed keyboard with chording for more keys) and a mouse. It was perfect for WoW which required something like 50 different keys to be pressed with one ha
20.
by alasdair_ 2mo ago
It knows when several tokens are about equal vs. times where one token is vastly preferred. In the latter case, that’s usually code or math or something similar and so it won’t alter those tokens.
21.
by alasdair_ 2mo ago
They say in the paper that they will not do this for code but it may happen for comments.
22.
by alasdair_ 2mo ago
That honestly seems reasonable. AI is more relevant than oil.
23.
by alasdair_ 2mo ago
The average family will be paying plenty for ai, just not directly in the form of tokens.
24.
by alasdair_ 2mo ago
Every llm-wrapper business needs an llm behind it. There are a sizeable number of firms paying 10mm+ a month to offset the people paying $50/month
25.
by alasdair_ 2mo ago
Enterprises will pay tens of millions a month, millions of individuals will pay $100 a month and there will be a long tail as they offer cheaper pricing and perhaps ad-supported pricing. The average American family won’t be willing to pay m
26.
by alasdair_ 2mo ago
What would a hedge fund want to do on this exactly? It’s too slow for hft and I’m not sure what they would be doing where ms matter but is not hft.
27.
by alasdair_ 2mo ago
Oh it’s worse than that. There are places where you CAN set a limit. This seems great until you are at the center of a huge traffic spike because of good pr and so you try to change it to a larger number only to be told you need to wait 24
28.
by alasdair_ 4mo ago
I wrote a rules engine in rust along with a reinforcement learning with MCTS based system to play decks against each other. It can handle aggro decks well enough but complex combo decks like Amulet Titan are tough to get working without exp
29.
by alasdair_ 4mo ago
Does this mean Mamdani can build a datacenter in the middle of central park if he wanted to?
30.
by alasdair_ 4mo ago
If the scales are only checked after the heat death of the universe, does it even matter? If the market can’t actually detect crooks and charlatans until long after they have stolen investors money, its ability to be “correct” is worthless.

More ›