dpflan

50 ms·

571.
OpenAI losses doubled to $540M last year as revenue quadrupled(twitter.com)
23 points by dpflan 3y ago|3 comments
572.
by dpflan 3y ago
Completely agree. It is interesting, but the gravitas of it seems lower than of course if an executive said this and corroborated it. I do feel that opensource for AI is going to be really interesting and shake things up.
573.
by dpflan 3y ago
Yes, these huge models are the expert systems for training specialized systems that will run often and need to run cheaply.
574.
by dpflan 3y ago
Finetuning models and LLMs (and any model) is going to a be common practice . Each company is its own domain, which domain knowledge and data to specialize open sourced models or used other models to distill/teach their own proprietary
575.
by dpflan 3y ago
Could LLMs help here to translate between programming languages? Same with COBOL. Is a coding assistant the way to get legacy systems updated -- sounds like a consulting niche.
576.
by dpflan 3y ago
This may seem silly but can there be a text transformation/translation layer, where the conversation text is ultimately PG, the user thinks they have sent their thoughts, but they are translated and vice versa for the user and AI, so i
577.
by dpflan 3y ago
Consolidate with this one? https://news.ycombinator.com/item?id=35770048 it has discussion already. (395pts / 324 comments) vs. (28pts / ~0 comments)
578.
by dpflan 3y ago
I wonder if the compute power/GPUs for crypto mining are being converted to be compute for LLMs/GenAI/AI. I wonder because I wonder what percent of crypto compute resources that are under the custodianship of "bad actors
579.
GenPhys: From Physical Processes to Generative Models(arxiv.org)
1 points by dpflan 3y ago|0 comments
580.
by dpflan 3y ago
Ah, so being outside in nature...
581.
by dpflan 3y ago
Check this out: https://news.ycombinator.com/item?id=34963057 From a previous article discussion on the vertical farming bubble popping: https://news.ycombinator.com/item?id=34959649 """ It a
582.
by dpflan 3y ago
I mean, this is pretty incredible: https://research.facebook.com/blog/2023/4/every-tree-counts-...
583.
by dpflan 3y ago
Didn't personal computers in general began as toys for hobbyists? Look where that got us...
584.
Geometry and visualizations of linear programs [pdf](ocw.mit.edu)
2 points by dpflan 3y ago|0 comments
585.
by dpflan 4y ago
I think I was more thinking about the possible direct mapping of the physical device to the computational device implying that it may not be possible to make make a more intelligent device from a device base. What is your lab doing? Are you
586.
by dpflan 4y ago
FTA: "A quintessential cognitive task used to measure human working memory is the n-back task. In this study, task variations inspired by the n-back task are implemented in a NWN device, and external feedback is applied to emulate brai
587.
by dpflan 4y ago
GAT is particularly interesting at the moment: https://scholar.google.co.uk/citations?view_op=view_citation...
588.
by dpflan 4y ago
- https://geometricdeeplearning.com/ -- group of researchers presenting the paradigm - https://petar-v.com/ - Petar's homepage - https://scholar.google.co.uk/citations?user=kcTK_FAAAAAJ
589.
by dpflan 4y ago
Interesting. Could one graph be the topological representation, let's say like SCC? So there are heterogenous graphs with meta-graphs connected to them...?
590.
by dpflan 4y ago
Thank you. "Reading the first part of the present paper, the Topological DL would instead move beyond the idea of "pairwise relation". -- interesting. It seems to make sense intuitively, given difficulties in GNNs, to look at
591.
by dpflan 4y ago
I am not well versed in topological deep learning; but can topological concepts be learned by a graph neural network -- like can there a be topological layer(s) that learns the topology of the graphs, like how it's easy to visualize ho
592.
by dpflan 4y ago
FTA: But what sets Community apart is the dialogue that celebrities and brands have with their customers, who provide troves of information about themselves, which the brand owns and isn’t shared with Community’s other clients. Oseary was o
593.
by dpflan 4y ago
sure
594.
by dpflan 4y ago
Why isn't this called "prompt-imization"? Maybe prompt-opt or opt-prompt is easier...
595.
by dpflan 4y ago
It is certainly an interesting phenomenon, and I wonder what techniques from neuroscience for brain mapping could be used for model "brain mapping", which could lend itself more to prompt engineering as a science (latent space map
596.
by dpflan 4y ago
Interesting project…thanks for sharing.
597.
by dpflan 4y ago
A glaring problem is: Non-determinism of LLMs, creating different answers to the same prompt. I appreciate your blogging and analysis in this space, so I am I am interested in your responses. The non-determinism implies that prompt engineer
598.
by dpflan 4y ago
Is it useful in other legal cases? If you can reference this case in your own case and evidence of patterns of behavior? IANAL
599.
by dpflan 4y ago
It's a money making scheme for sure, with gatekeepers and entrenched agents, not all medical providers are ethical...
600.
by dpflan 4y ago
The narcissism pyramid concentrates its effects at the top. We live in world now where narcissism and destructive personalities can be reinforced, especially so via social media. There have always been these issues with humans interacting a

More ›