HNHacker News
TopNewBestAskShowJobs

saliagato

380 karma · joined May 31, 2022

submissionscomments
saliagato··on Coinbase awarded a $500k bug bounty
Can't comment on the cost (varies depending on who you are and what you do in your life) but the benefit is really high.

source: I started from 0 and the very first bounty I submitted (related to CEX) was valued to $100k+

saliagato··on The "full-stack framework" fallacy
Speed of development, developer experience, ready to go templates. You should check it out, it’s really cool
saliagato··on GPT-4o
gpt-4 was indeed trained on gpt-3 instruct series (davinci, specifically). gpt-4 was never a newly trained model
saliagato··on Show HN: Beyond text splitting – improved file parsing for LLMs
worked 100% of the time for me
saliagato··on The Terrifying A.I. Scam That Uses Your Loved One's Voice
This sounds so fictionalized that may be fake
saliagato··on Google to Pause Gemini Image Generation of People After Issues
This post is still the first in the feed
saliagato··on Our next-generation model: Gemini 1.5
basically, yes. Pinecone? Dead. Azure AI Search? Dead. Quadrant? Dead.
saliagato··on Qdrant, the Vector Search Database, raised $28M in a Series A round
Hard to believe OpenAI uses Quadrant when they are backed by Microsoft, thus having Azure Cognitive Search (now "AI" Search)
saliagato··on ChatGPT for Teams
no
saliagato··on Epstein documents: Names cited in court papers are released
I wonder how many were left out
saliagato··on Mixtral of experts
Same. Tip: if you put a random number it will get you through regardless
saliagato··on Stripe live dashboard
this is so cool lol
saliagato··on We have reached an agreement in principle for Sam to return to OpenAI as CEO
Why do you (dang) always write a comment specifying that people can read more and even providing some links when it's clear that when you reach the bottom of the page you have to click "read more" to indeed read more. Isn't it a bit useless?
saliagato··on Microsoft: We remain committed to our partnership with OpenAI
same topic different tweets
saliagato··on OpenAI's board has fired Sam Altman
Hopefully they don’t
saliagato··on OpenAI's board has fired Sam Altman
We all know what. HN moderators are deleting all related comments.

Edit: dang is right, sorry y’all

saliagato··on Exploring GPTs: ChatGPT in a trench coat?
I think they are running their RAG on Microsoft's Azure Cognitive Search. Satya mentioned something about it in the yesterday Microsoft Ignite keynote
saliagato··on Is the reversal curse in LLMs real?
It can’t count because what the LLM sees is a bunch of tokens not words
saliagato··on Ask HN: Is ChatGPT4 broken since last weekend?
Yeah same. I switched to Azure a while back but it's unusable now. Switched back to OAI API
saliagato··on OpenAI Data Partnerships
Basically they take your data for free and they will receive your money from the API usage you will make.

Good

saliagato··on New models and developer products
You can now [1] pay from $2 to $3 million to pretrain custom gpt-n model. This has gone unnoticed but seems really neat. Provided that a start-up has enough money spend on that, it would certainly give competitive advantage.

[1] https://openai.com/form/custom-models

Edit: forgot to put the link

saliagato··on New models and developer products
I think your startup just died
saliagato··on New models and developer products
sama said they improved it
saliagato··on OpenAI’s November 6th DevDay has been leaked
OpenAI wants to make ChatGPT an ecosystem similar to the App Store. They are creating a diversification in their income: -Subscriptions -API -This new marketplace

That’s a nice strategy

saliagato··on How to Not Get Screwed over as a Software Engineer [video]
Care to elaborate?
saliagato··on Nobel Prize in Physics Awarded to Agostini, Krausz, and L’Huillier
which?
saliagato··on Satya Nadella Says Copilot Will Be as Significant as the PC
yes
saliagato··on LeoLM: German-Language LLM Research
OP was referring to nazis
saliagato··on Amazon will invest up to $4B in Anthropic
Europe
saliagato··on GPT-4 is not getting worse
I can get 8k + tokens response from OpenAI API without problems (with gpt-3.5-turbo-16k) but I encounter errors from Azure OpenAI when using the same prompt and parameters. Obviously it sucks because I can't get long response from gpt-4-32k on Azure
Page 1 of 3Next →