sandkoan

7 ms·

31.
by sandkoan 3y ago
Shoot me an email at govind <at> automorphic <dot> ai
32.
$100 If You Can Beat Our Firewall and Trick Our Model(automorphic.ai)
2 points by sandkoan 3y ago|0 comments
33.
by sandkoan 3y ago
Would love an invite as well, if you still have some :)
34.
Anything Can Be a Weekend Hack(blog.govindgnana.com)
2 points by sandkoan 3y ago|0 comments
35.
by sandkoan 3y ago
Wholeheartedly agree—it adds layers of bloat and abstraction between you and the actual ReAct pattern, which can be trivially implemented in maybe 50 sloc.
36.
by sandkoan 3y ago
Do you have a copy of the script you used to export all your notes? I'm currently faced with this existential problem myself.
37.
by sandkoan 3y ago
"RTFM" means "Read The Fucking Manual."
38.
by sandkoan 3y ago
Yeah, users can set their own list of redacted words.
39.
by sandkoan 3y ago
Good idea! Adding that as we speak.
40.
Show HN: Firewall for LLMs–Guard Against Prompt Injection, PII Leakage, Toxicity(automorphic.ai)
17 points by sandkoan 3y ago|7 comments
41.
Databricks acquires OpenAI Competitor MosaicML for 1.3B(mosaicml.com)
3 points by sandkoan 3y ago|0 comments
42.
by sandkoan 3y ago
You're absolutely right. The comedic relief scene is almost necessary to avoid us being stretched too taut after the previous. Thomas de Quincey has a great piece reflecting on the very same: http://www.shakespeare-online.co
43.
by sandkoan 3y ago
Something which may not be known to many, McCarthy was also quite interested in math, philosophy, cognition. A piece of his non-fiction work which made a lasting impression on me was "The Kekule Problem" [0], exploring the power o
44.
by sandkoan 3y ago
Presumably, they're referring to chunking up the data into discrete semantic units—smaller vectorizable subsections (e.g., paragraphs) more precisely capturing different parts of the data.
45.
by sandkoan 3y ago
Isn't it? Wrangling a poorly understood system with partial knowledge and experimental skill seems like engineering to me.
46.
Inflection – A Personal AI(inflection.ai)
1 points by sandkoan 3y ago|0 comments
47.
by sandkoan 3y ago
The imagery you just evoked reminds me of Bradbury's short story, "The Veldt." [0] [0]: https://repositorio.ufsc.br/bitstream/handle/123456789/16372...
48.
by sandkoan 3y ago
https://oa.mg/pdf/10.1177-0956797620903113.pdf
49.
by sandkoan 3y ago
I recall Ben Horowitz saying a16z was a conscious nod to i18n in some interviews.
50.
by sandkoan 3y ago
Bubeck has clarified that the "1 trillion" number he was throwing around was just a hypothetical metaphorical—it was in no way shape or form implying that GPT-4 has 1 trillion parameters [0]. [0] https://twitter.com&#x2
51.
by sandkoan 3y ago
I'm curious as to what people do currently, without runhouse. And how does runhouse work, under the hood? Could I trivially configure model serving without resorting to FastAPI or some such?
52.
by sandkoan 3y ago
When you say you wish they had a "better story for local development," what do you mean? What do you wish for?
53.
by sandkoan 4y ago
You may want to look at run.house [0] for a pretty powerful solution to many of these problems. [0] https://github.com/run-house/runhouse
54.
by sandkoan 4y ago
I only brought up the Gladwell post to clarify the scale at which these institutions operate. There is no question that they have the money, and, if they chose, could sustainably provide scholarships to some subset of the particularly mer
55.
by sandkoan 4y ago
No, I'm criticizing their decision to only offer need-based scholarships when they could just as well provide both—if they really cared about academic strength, then they could put their money where their mouth is, and provide meaningf
56.
by sandkoan 4y ago
The Ivies and adjacent elite colleges are more than capable of giving out academic scholarships—they simply refuse to because that would lead to a race to the bottom in terms of how much money they'd be able to extract from top talent
57.
by sandkoan 4y ago
I'm actually building this very thing—shoot me an email at govind <dot> gnanakumar <at> outlook <dot> com if you'd like to be a beta tester.
58.
by sandkoan 4y ago
I'm actually building this very thing—shoot me an email at govind <dot> gnanakumar <at> outlook <dot> com if you'd like to be a beta tester.
59.
by sandkoan 4y ago
Presumably, because they trained Copilot on billions of lines of, often licensed, code (without permission), that Copilot has a tendency to regurgitate verbatim, without said license.
60.
by sandkoan 4y ago
Might this be what you're looking for: https://github.com/bigscience-workshop/petals ?

More ›