HackerTrans
トップ新着トレンドコメント過去質問紹介求人

VectorVault

no profile record

投稿

How are you governing token spend and data exposure in prod AI agent loops?

news.ycombinator.com
2 ポイント·投稿者 VectorVault·3 か月前·0 コメント

コメント

VectorVault
·3 か月前·議論
[dead]
VectorVault
·3 か月前·議論
Most politics now smacks of satire and performative outrage. "I can solve the problems I created with more of someone else's money..."
VectorVault
·3 か月前·議論
Snow Crash came to mind, but so did several other Neal Stephenson books!
VectorVault
·3 か月前·議論
Scary. I'm not familiar with the work. Though, it strikes me as a lot like a Neal Stephenson novel unfolding in our faces.
VectorVault
·3 か月前·議論
[dead]
VectorVault
·3 か月前·議論
Can you see a world where everyone has an AI Persona based on their prior work that acts like a RAG to inform how things should be coded? Meaning this is patent qualified code because, despite being AI configured, it is based on my history of coding?
VectorVault
·3 か月前·議論
What's really tricky with the legal protections area is this: 90% of the value of the S&P 500 is intangible. Meaning if you suck out the book value (10%), the rest is brand, IP, rights, sources & methods, etc. So if a company can't protect that, it's not particularly valuable anymore. Maybe we will see a shift back to tangible assets and book value (25,000 $8MM Vera Rubin machines) and away from intangibles...
VectorVault
·3 か月前·議論
Hmm... It's an interesting question: at what point, in a conversation, document, image, story, does human authorship end and AI authorship begin? How would you know? Is it a tell tale or is it advertising?
VectorVault
·3 か月前·議論
Strict grammar school teachers, who enforced a "Form And Style" level of prose, have become a liability in the AI world!
VectorVault
·3 か月前·議論
It's interesting how AI can be its own worst enemy in this legal system. The very thing it's excellent at is not protected. In practice, there seems to be a strong opportunity to disintermediate brands by acting as a layer of abstraction above the seller and manufacturer. An AI instruction likely cares less about brand or sharing customer information with the seller; it's just more friction and tokens spent.