localhost

5 ms·

61.
by localhost 3y ago
Many of the research grants are funded from public tax dollars
62.
by localhost 3y ago
Porsche for one - don't know of others.
63.
by localhost 3y ago
It's under experimental features or something like that. I just enabled Bard access for myself yesterday.
64.
by localhost 3y ago
I'd love to have something that could create daily summaries of private Twitter lists that I use to track developments in AI. The audio part would be great to consume during my re-emerging commute to the office. I would imagine this wo
65.
by localhost 3y ago
I asked GPT-4 to explain it to me, and asked if it could identify the algorithm. It didn't do a good job here (returned name of function). But when I told it that it was also called "Shar's algorithm", it immediately ide
66.
Porsche Synthetic Fuel in Chile(youtube.com)
1 points by localhost 3y ago|0 comments
67.
by localhost 3y ago
While not a fix for this problem, a nice form of insurance is a Lithium Ion jump start battery[1]. I'm amazed that a hand-held battery can discharge fast enough to jump start my car, but ... it works great. [1] https://www.a
68.
by localhost 3y ago
While I don't remember to use it (too much AI in too many tools!) Warp has this feature as well "Warp AI". This is what it told me for the demo scenario: ffmpeg -i input.mp4 -i watermark.png -filter_complex "overlay=
69.
by localhost 3y ago
I did an experiment where I got ChatGPT to rewrite headlines in neutral tone based on the content of the article. Works great. What you want is a great idea!
70.
by localhost 4y ago
I copied the response verbatim. I also did a follow-up where it wrote the interpreter as well: class KimmieInterpreter: def __init__(self): self.variables = {} def interpret(self, code):
71.
by localhost 4y ago
Since I have the world's greatest creative assistant handy, here's what GPT-4 thinks a Kim Kardashian language would look like: Hey dolls, let me introduce you to the Kimmie programming language, it's like totally fab and eas
72.
by localhost 4y ago
This is how few-shot prompting works and it dramatically improves the quality of the results.
73.
by localhost 4y ago
This is a good idea that you could likely build out using just embeddings and performing semantic search over them. ChatGPT could be relegated to a role of translating intent into semantic searches and refining the output. You could even im
74.
by localhost 4y ago
I think it would be quite a bit more interesting if you could converse with the model. The back and forth "is this paper about foo related to this other paper about bar?" would probably be a better way of getting at the interconne
75.
by localhost 4y ago
There's a guy on YouTube (wfh - Work From Hype) who has a channel devoted to these kinds of things: https://www.youtube.com/channel/UCcCYTCDxsMl84_btikjY40Q
76.
by localhost 4y ago
This is a much better link than the GH repo - it contains examples and the docs. https://motioncanvas.io/
77.
by localhost 4y ago
If you change the goal from "I wrote a book" to "I published a book and made money off of it", it's really no different than fronting the money to pay for people to write and illustrate the book, right? I tend to th
78.
by localhost 4y ago
I'm a big fan of Lilex and use it currently: https://github.com/mishamyrt/Lilex . I think it looks quite a bit like Inconsolata which I used to use a long time ago.
79.
by localhost 4y ago
Relevant to this discussion is the Toolformer paper that was just released today from Meta AI. To me the exciting result from the paper is that it appears that much smaller models can be used to translate human intent into actions: https:&
80.
by localhost 4y ago
The key to getting good results with Large Language Models is to combine them with another tool. For example, a human with a search engine produces more accurate results than a human alone. In this spirit, I chained together GPT 3.5 and a s
81.
by localhost 4y ago
I suspect that there is going to be a push to the extremes in dev work. You will either need to be effective as a generalist who uses AI powered tools to rapidly build systems out of current libraries, or you need to be a specialist who mai
82.
by localhost 4y ago
GPT3 is ~175B parameters. At float16 precision, that's 350GB of weights. BLOOM-176B is about the same size. Here's one person's experience; a token is ~0.75 words. "The Python code in this tutorial generates one token ev
83.
by localhost 4y ago
The attention mechanism in transformer models also uses this trick - just dot product. I learned this during a long conversation with ChatGPT the other night - an excerpt of its reply to me: The dot product between the query and key rep
84.
by localhost 4y ago
I was playing around doing the same thing today but using Pinecone [1] as a vector store. Some other folks were experimenting with building key-value stores at the Scale AI hackathon this weekend [2] [1] https://www.pinecone.io&#
85.
by localhost 4y ago
The author of this paper runs the inference team at OpenAI. So I would imagine that this is a record of her experiments without necessarily revealing which ones they're actually using in production.
86.
by localhost 4y ago
But wouldn't you need to integrate Python into Excel for this to work?
87.
by localhost 4y ago
Humans tend to not like change, particularly in institutions that are themselves resistant to change. Education today is very much like the education in the early 1900s - designed to educate a compliant workforce who would work in factories
88.
by localhost 4y ago
This tweet has a lot of additional details with timelines for releasing weights of LLMs: https://twitter.com/swyx/status/1579551634832883712?s=20
89.
by localhost 4y ago
Not that I know of. But my understanding is that's their intent.
90.
by localhost 4y ago
I don't have an opinion on this one way or another, but if the RAIL license concerns you, then perhaps you can take it up with the organization behind it? https://www.licenses.ai/

More ›