curious_cat_163

14 ms·

91.
by curious_cat_163 2y ago
QuietBox is $15,000 not $1,500: https://tenstorrent.com/hardware/tt-quietbox
92.
by curious_cat_163 2y ago
Talk about demand generation!
93.
by curious_cat_163 2y ago
Not trust Microsoft to be the operating system of the modern life?
94.
by curious_cat_163 2y ago
I took this line a bunch about two weeks ago without realizing its history. Tokyo’s public transit system is the best I have ever seen!
95.
by curious_cat_163 2y ago
Articles like this represent reminder that we are in the middle of a hype cycle. [1] I don’t say that thinking that LLMs (really: Transformers and the corresponding scaling of compute around it) don’t represent a step change. I say that be
96.
Do we need language to think?(nytimes.com)
3 points by curious_cat_163 2y ago|1 comments
97.
by curious_cat_163 2y ago
It’s a pretty clever idea: “check” if the model answers “differently” when asked the same question again and again and again. “checking” is being done with another model. “differently” is being measured with entropy.
98.
by curious_cat_163 2y ago
Thanks for sharing your insights. One wonders if the business model for supporting efforts like this could be to provide the support/services, while keeping the code base open source?
99.
by curious_cat_163 2y ago
I have personally found them useful as “glue” layer in heterogeneous and large systems with moving and decoupled parts. There are any number of ways to do the same thing — context matters.
100.
by curious_cat_163 2y ago
I have yet to find a use case where quality can be traded off. Would love to hear what you had in mind.
101.
by curious_cat_163 2y ago
I agree: hybrid teams with a diversity of product/research skills at the team level is the way to go. It is thinkers and doers that need to come together. It is way easier said than done, though. You need true buy in from a ton of stak
102.
by curious_cat_163 2y ago
> While no one is getting as much computing power as they want, the supply is tighter for teams engaged in pure research, say the former employee and others familiar with the lab. Good! Maybe they will focus on researching how to make th
103.
by curious_cat_163 2y ago
> I built ThreadQuilt because I was tired of wading through endless noise to find meaningful discussions. With ThreadQuilt, you get a cleaner, more focused experience that highlights the best parts of the web's conversations. I appl
104.
by curious_cat_163 2y ago
So, this is a good idea. Having opinions about what AGI benchmarks should look like is a great way to argue about the kind of technology we want to build for the future. However, why are the 100 test tasks secret? I don't understand wh
105.
by curious_cat_163 2y ago
No.
106.
by curious_cat_163 2y ago
I agree with you about this being a bad precedent. However, to me, the off-device bit they showed today (user consent on every request) represents a strategic hedge as a $3T company. They are likely buying time and trying to prevent people
107.
by curious_cat_163 2y ago
Reminds me of FastUI: https://github.com/pydantic/FastUI
108.
by curious_cat_163 2y ago
Cute! The signal processing folks have entered the room... :)
109.
by curious_cat_163 2y ago
> a stateful load balancer that is aware of each server's available slots Interesting. Curious to understand what a 'slot' is, in this context. Is that a llama.cpp specific application-layer state that llama.cpp makes avai
110.
How Americans View Data Privacy? (Pew)(pewresearch.org)
3 points by curious_cat_163 2y ago|0 comments
111.
by curious_cat_163 2y ago
> Beyond the technology community I doubt that the loss of personal privacy is being discussed at large. Actually, people in the US are worried about privacy in the age of AI. [1] [1] https://www.pewresearch.org/internet&
112.
by curious_cat_163 2y ago
A paragraph ends with: > This could arguably be seen as a violation of the Cloudflare TOS, as they wrote above. And the very next paragraph begins with: > In any case, we receive >95% of our traffic through the main domain that’s b
113.
by curious_cat_163 2y ago
Okay. Respect your point of view. I am curious, what applications do you think SSMs enable that a Transformer cannot? I have always seen it as a drop-in replacement (like for like) but maybe there is more to it. Personally, I think going li
114.
by curious_cat_163 2y ago
IMO, SSMs are an optimization. They don't represent enough of a fundamental departure from the kinds of things Transformers can _do_. So, while I like the idea of saving on the energy costs, I speculate that such saving can be obtained
115.
by curious_cat_163 2y ago
There is a tick-tock between searching the dominant NN architectures (tick) and optimizing for accuracy, compute and inference latency and throughput (tock). This particular (tock) is still playing out. The next (tick) does not feel imminen
116.
by curious_cat_163 2y ago
The nerve.
117.
by curious_cat_163 2y ago
Cool! All the best with the build and launch. FWIW, I think that the conversion would be fine either way. If you haven’t already, it might be worth adding a free tier. The data from the free tier then could answer the questions around what
118.
by curious_cat_163 2y ago
I wonder if any generalization as an attribute for all VCs is ever going to be a good enough. People call them “not friends” or “sheep” or whatever. But it has always felt like an unsatisfactory framing. It is a marketplace. The equity of c
119.
by curious_cat_163 2y ago
Put another way: the US investors put a lot of net new capital every month. Theoretically, more capital means, higher demand for equities and hence better prices for the stock. Of course, this does not apply to all equities equally.
120.
AI Economics(benjaminfspector.com)
1 points by curious_cat_163 2y ago|0 comments

More ›