davesque

12 ms·

151.
by davesque 2y ago
Maybe he's just facing the reality of polling data and not concealing a health problem?
152.
by davesque 2y ago
As a person with a music degree, albeit not in composition, but I had several friends who were composition majors, my impression is that composers often do write music in old baroque or classical styles etc., which are academically well u
153.
by davesque 2y ago
Also seems important to ask why these patterns became so common and the related question of why companies are so much more willing to violate people's dignity nowadays just to get an edge on the competition. Was it the rise of the MBA?
154.
by davesque 2y ago
My intuition is that there must be some relatively simple ways to tweak the format of assignments or classes so that it's harder to use tools like ChatGPT to cheat.
155.
by davesque 2y ago
Then why would you choose to use a CNN over a fully connected neural network for visual data? For the same reason you'd choose a KAN over a traditional neural network if you were trying to fit a continuous function that can easily be m
156.
by davesque 2y ago
A case study in hype and FOMO.
157.
by davesque 2y ago
KANs seem like a great tool for the right job. However, based on my understanding of how they work, my intuitions tell me that they would be awful at image processing, which I think was one of the author's test beds.
158.
by davesque 2y ago
Worth repeating, though.
159.
by davesque 2y ago
> given your incredibly thinly veiled bias What bias is that?
160.
by davesque 2y ago
I flagged it because it's really boring to see people go through the same predictable, tribal motions on a topic that's been covered to death by every media outlet.
161.
by davesque 2y ago
I always felt like the one thing I wanted in a monad tutorial was a list of things that aren't monads because they follow all the monad rules except for one, for each rule.
162.
by davesque 2y ago
If the parquet file includes any row group stats, then I imagine DuckDB might be able to use those to avoid scanning the entire file. It's definitely possible to request specific sections of a blob stored in S3. But I'm not fami
163.
by davesque 2y ago
This echoes my feelings on this whole debate. The question of consciousness matters much less than the question of how we decide to treat things that we view as being very much like us.
164.
by davesque 2y ago
Honestly seems like you could also easily just laugh it off. People on the internet are weird and gross. But I guess it sounds like he was already kind of over the project anyway.
165.
by davesque 2y ago
Though this article does point out what I think is a real effect in human psychology, it bothers me that the author ends up using the whole narrative to inject her own opinions on world affairs in a context where we should feel guilty for q
166.
by davesque 3y ago
I visited Hualien a few years ago. Very sad to see this.
167.
by davesque 3y ago
Something I've always noticed with the notion of Kolmogorov complexity is that the question of determining the lowest level of computation is problematic. For example, in the article, the author first defines the basic idea of KC. But
168.
by davesque 3y ago
I feel like this article should be called "The Law of Bad Abstractions." I often see this cited as a blanket rejection of complexity in software. But complexity is unavoidable and even necessary. A skillful engineer will theref
169.
by davesque 3y ago
Prove it.
170.
by davesque 3y ago
Seems spammy to drop a 100k line PR of just comments, AI generated or otherwise. No maintainer in their right mind would want to review that.
171.
by davesque 3y ago
Not necessarily. Companies can be rudderless and have some kind of industry capture or brand recognition and still be doing great. In other words, at one point they had good leadership or built up during times of low regulation or economi
172.
by davesque 3y ago
For the sake of argument, does that mean all those employees were useless or does it mean the management isn't doing a good job of allocating them to effective work? I've been at a few companies that were rudderless and yeah, lett
173.
by davesque 3y ago
On the bright side, it might force those engaged in research to figure out just what exactly it is that makes their models work to begin with. Turns out knowledge is about more than just modeling statistics.
174.
by davesque 3y ago
> Lava has flowed over Njarðvíkuræð, the hot water pipeline that runs from HS Orka's geothermal power plant in Svartsengi to Njarðvíkur, causing supply issues. I would think so lol.
175.
by davesque 3y ago
It's only technically better that it was a quote. That doesn't come close to excusing the behavior. It's still completely beyond that pale that he chose those specific lyrics to post on Twitter. The question of how honest
176.
by davesque 3y ago
I don't see how it makes it any better that it was a quote. The quote still had words in it. Does Garry Tan think the entire world listens to Tupac? And what wisdom does Tupac really have to offer in this case?
177.
by davesque 3y ago
Yeah, I guess you're right. The queue vs. addressable memory thing seems secondary. Is the real difference that a transformer model is stateless and Turing machine is stateful? And the assumption of a stateless model is why they can
178.
by davesque 3y ago
Yeah, I sort of agree. It hardly provides any details about the model that they supposedly used to achieve only two benchmark results which are poorly described. It comes across as being written by someone that only dabbles in the topic b
179.
by davesque 3y ago
There's a PDF link next to the title on the linked page, but here's a direct link: https://openreview.net/pdf?id=MGWsPGogLH
180.
by davesque 3y ago
Okay, actually I think I just misunderstood the argument in the paper. After taking a second look, I think their argument goes as follows: A typical autoregressive transformer model operates on a fixed-size context that acts as both input

More ›