bartuu

8 ms·

1.
by bartuu 2y ago
It's nearly impossible to prevent LLMs from hallucinating, which creates a significant reliability problem. Enterprise companies think, "Using agents could save me money, but if they do the job wrong, the damage outweighs the bene
2.
Show HN: Reliability layer to prevent LLM hallucinations(github.com)
3 points by bartuu 2y ago|1 comments
3.
by bartuu 2y ago
but server-client architecture makes debugging more challenging. How do you handle debugging in distributed environments?
4.
by bartuu 2y ago
Generating documentation with different agents, each in their own image? Yeah, been there. A shared base image would've definitely simplified things.
5.
by bartuu 2y ago
I agree. The other day I went to an event. One of the speakers said he had a $2.7 billion exit. Everyone in the room believed it because some brilliant people and high-level authorities in the room believed in him and put him on that stage.
6.
by bartuu 2y ago
I get excited when I see things like this, even if they're simple because I think I can build a business on top of it. However, in complex tasks and real-life cases, it's successful in very few instances. I can't trust its st
7.
Building Effective Agents(anthropic.com)
9 points by bartuu 2y ago|0 comments
8.
by bartuu 2y ago
I wonder, if Apple made a deal with Open AI, how did they solve the privacy issue?
9.
by bartuu 2y ago
Hi im Onur, As someone with over five years of experience in Python serialization and the Django framework, I've been heavily involved in the development and maintenance of Tiger, a function store that provides isolated storage, auto-d
10.
by bartuu 2y ago
As someone with over five years of experience in Python serialization and the Django framework, I've been heavily involved in the development and maintenance of Tiger, a function store that provides isolated storage, auto-documentation
11.
Transform LLMs into real-world operations(github.com)
1 points by bartuu 2y ago|1 comments
12.
by bartuu 2y ago
Hi, i am Onur Ulusoy, a Python enthusiast whose experience spans over 5 years with a special focus on Python Serialization over the past 2 years. In start we just working on fastest Deployment system in python (<1) but after a pivot in U