dekhn

6 ms·

91.
by dekhn 2mo ago
You are not being prevented from accessing another copy of the book to get that knowledge.
92.
by dekhn 3mo ago
These articles are designed to create maximal outrage, and they can mislead you into thinking the problem is far worse than it really is. Conflating the nature of the books (between "rare" and "common") is exactly the s
93.
by dekhn 3mo ago
I do too. But my hope is that by calmly repeating the truth the AI haters might realize that they are being manipulated.
94.
by dekhn 3mo ago
This is one of those articles that goes out of its way to enrage people, when the underlying reality is far simpler, more straightforward, and less nefarious.
95.
by dekhn 3mo ago
It's hard to take the article seriously when it has quotes like this: "The hardest part. 2B edges twitter graph is already huge (its edges are 30 GB in CSV !!!)." Who cares how big the graph is in CSV? That's not the re
96.
by dekhn 3mo ago
Got it. I guess there is a whole subgenre of 8-bit pixel art tattoos as well.
97.
by dekhn 3mo ago
Interesting. I have noticed there are a few spots in my body, where I feel sensation far from the origin (for example, my finger twinging slightly when I pee). I always wondered about the nerve mappings in that case.
98.
by dekhn 3mo ago
Was this meant seriously? I literally used those graphics 40 years ago (it's from a wonderful program called PrintShop ( https://news.ycombinator.com/item?id=41087798 )
99.
by dekhn 3mo ago
Are tattoos normally considered painful? I got my first one recently (a small shoulder tattoo) and it didn't hurt really, more like a cat's tongue rasping against the skin. I have heard my artist has a light touch, and it wasn&#
100.
by dekhn 3mo ago
I work in pharma and a coworker asked if they could use my code (a model inference server). I asked what they were doing and it's an ML model for determining clinical dosing for a cancer trial. They sequence a person's cancer, t
101.
by dekhn 3mo ago
P = NP is actually one of the worst abuses of symbology I've seen in math. """During his own Google interview, Jeff Dean was asked the implications if P=NP were true. He said "P = 0 or N = 1." Then, before the
102.
by dekhn 3mo ago
I would expand on this. AI is great for me because it can read the equations I don't understand and turn it into code I can understand. I've worked in science for decades and it's still like pulling teeth to replicate a com
103.
by dekhn 3mo ago
I had a few moments of this in the past. For example, in my quantum class the teacher wrote "H Psi = E Psi" on the board, we all laughed, "just cancel the psi" but it turns out one was a multiplcation and the other was
104.
by dekhn 3mo ago
I mean, if you make a paperclip optimizer, don't be surprised if it optimizes your paperclips.
105.
by dekhn 3mo ago
He stopped by my booth at Open Sauce yesterday, I let him know his work is awesome, inspiring, and also causes a little imposter syndrome. I'm just glad he shares his results with us!
106.
by dekhn 3mo ago
I love how "safe mode" for a satellite is basically: "extend solar panels, turn self towards sun, don't do anything unnecessary, wait for further instructions".
107.
by dekhn 3mo ago
Practically speaking, while we could simulate them at a fairly approximate level, it wouldn't really tell us anything useful.
108.
by dekhn 3mo ago
Good point. I've ended interviews (as a candidate) when I believed the interviewer was being lazy (for example, administering a leetcode question and not even changing any of the details like input or output data). I ended up writing
109.
by dekhn 3mo ago
Literally the second I read "it's a data stream" I knew the answer was going to be reservoir sampling. RS is really interesting to me. many people you talk to can realize you can compute the mean of a data stream ( https:&#x
110.
by dekhn 3mo ago
Did you actually find any actionable results? When I had my genome sequenced professionally they told me I literally had no SNPs in my genome that were associated with deleterious mutations. And even if you do, there are only some treatme
111.
by dekhn 3mo ago
.... which does not need or really want home sequencing. That's clinical-grade with QC and high quality instruments, along with professional genetic analysis.
112.
by dekhn 3mo ago
Nothing about this is the future. Sequencing at home will not solve any major problems. It's mainly a fun exercise to demonstrate that sequencing has been commodified.
113.
by dekhn 3mo ago
I'm of the general belief that something can be unaware and not have any understanding (in the subjective experience of consciousness), while also appearing to do so, and be useful. That's what Bender doesn't get. Some folks
114.
by dekhn 3mo ago
I mean, one of the most useful metal fabrication techniques is already folding thin sheets of metal; I learned it in shop class decades ago and it's still a very relevant skill. Some fun stuff here: https://www.youtube.com
115.
by dekhn 3mo ago
Yes, they are. Likely due to a deep relationship between math and physics, statistical modelling of complex natural phenomena has repeatedly been shown to be the most effective approach. This is true of LLMs, but also of many stochastic (
116.
by dekhn 3mo ago
There's plenty of beautiful math there, but the relationship to what our neocortex does is pretty distant. Individual biological neurons can do fairly complicated things, including compute 10-bit parity functions (you would normally
117.
by dekhn 3mo ago
Statistical models have repeatedly shown themselves to be the most productive research method for working with complex human-based systems (and in the larger study of natural phenomena). It remains unclear whether there is any short term p
118.
by dekhn 3mo ago
Personally, I've always read that paper as a political criticism of industry and industrialized research and capitalism. After decades in academic (and industrialized research) I've learned that smart people can write convincing
119.
by dekhn 4mo ago
Most of their fabs were divested to GlobalFoundries, but they still have pretty significant fab capability and capacity- I suspect at least partly to have a us-based chip-making for military ("Trusted Foundry").
120.
by dekhn 4mo ago
The author list will show "M.D." instead of (or in addition to) "Ph.D.". The paper terminology will be skewed towards medical terms.

More ›