shaburn

6 ms·

31.
by shaburn 3y ago
My primary question is from a legal standpoint, partially in response to new US regulation coming online.
32.
Ask HN: Who is responsible/liable for an LLM Output?
2 points by shaburn 3y ago|11 comments
33.
by shaburn 3y ago
GPT-4 with Web Browsing: Co-Founder vs LLMs General then Specific Situation... ____________ General... Advantages: Complementary Skills: Co-founder: 15 LLM/Agents: 0 Investor Confidence: Co-founder: 15 LLM/Agents: 0 Decision Makin
34.
by shaburn 3y ago
I am. And it is the best experience I've had so far, including having worked with technical founders from top schools and 8 figure exits.
35.
Ask HN: If you have LLM/Agents do you still need a co-founder?
4 points by shaburn 3y ago|4 comments
36.
by shaburn 3y ago
Only hundreds of billions at stake if successful.
37.
by shaburn 3y ago
Your Uber/AirBnB/Wework all have physical base units with ascending costs due to inflation and theoretical economies of scale. AI models have some GPU constraints but could easily reach a state where the cost to opperate falls and
38.
Ask HN: How are you sourcing your prompts for GPT?
1 points by shaburn 3y ago|0 comments
39.
by shaburn 3y ago
This will unlock a ton of Enterprise use cases. I have met with a number of enterprises that have a hard no on sending much of their data into OpenAI, even on Azure. I think this will persist as the brand faces many hurdles with respect for
40.
Ask HN: Is GPT-4 better at Bash or Python?
1 points by shaburn 3y ago|0 comments
41.
by shaburn 3y ago
Then what are the comp sci departments screening for and why?
42.
by shaburn 3y ago
No. And I beleive there is a ton of product to be built to account specifically for this in big systems.
43.
by shaburn 3y ago
I started on this 4 weeks ago on my own but haven't published.
44.
Ask HN: Why aren't we using ChatGPT in the CLI?
4 points by shaburn 3y ago|7 comments
45.
by shaburn 3y ago
Postgres and Mongo support it. ElasticSearch incorporates vector search. I have failed to identify an objective difference(beyond marketing nomeclature) between any pure play vector solution after exaustive research.
46.
Bloomberg Outlines Succession Plan for Bloomberg L.P(nytimes.com)
3 points by shaburn 3y ago|1 comments
47.
by shaburn 3y ago
https://archive.ph/oqhuJ#selection-353.0-353.53
48.
by shaburn 3y ago
That costs less than transportation to many doctors required for insurance approvals.
49.
Ask HN: Is there AI to close backdoors on your local system?
1 points by shaburn 3y ago|1 comments
50.
by shaburn 3y ago
Really curious how a profitable industry like medical research ranks against less profitible.
51.
by shaburn 3y ago
Speed and effort, possibly. Why use meld instead?
52.
Ask HN: How do you merge two files with ChatGPT(etc.)?
1 points by shaburn 3y ago|2 comments
53.
by shaburn 3y ago
Many bought crypto, few used it. Many use LLMs, few pay for them.
54.
by shaburn 3y ago
The state of California alone has 8 times your population and the US will gain your entire population in refugees in less than a decade.
55.
by shaburn 3y ago
I think Homeless is really a keyword for spending on <10,000 of the 750k+ people in SF who are X% target beneficiaries of those funds. Y% recipients dictate based on this <2% of the population.
56.
by shaburn 3y ago
There are 4 vacant apartments for every homeless person in San Francisco, including those in shelters. The idea of a net shortage is a farce.
57.
by shaburn 3y ago
Sorry.
58.
Salesforce Benioff warns he may pull conference from SF over homelessness drug(fortune.com)
29 points by shaburn 3y ago|34 comments
59.
by shaburn 3y ago
Why do you want to learn everything you do?
60.
by shaburn 3y ago
I'm building a local tool to read my system, files, and Git in sequence with ChatGPT API(and Gorilla and other LLMs soon)to avoid cut and paste threads by providing file data and specific situational(system) awareness.

More ›