Y
HN Search
Hacker News Search
new
|
comments
|
top
|
jobs
tzury
searching PlanetScale…
1.
▲
2.
▲
3.
▲
4.
▲
5.
▲
6.
▲
21 ms
·
481.
▲
Deciphering language processing in the human brain through LLM representations
(research.google)
3 points
by
tzury
2y ago
|
0 comments
482.
▲
Attention Is Logarithmic
(supaiku.com)
7 points
by
tzury
2y ago
|
0 comments
483.
▲
by
tzury
2y ago
$350M ARR in less than 5 years. Aiming towards $1B by the end of 2025. You never heard of them since perhaps your decisions were not in the cycles of their product. Those who are , heard indeed (type of folks who look at Gartner magic quadr
484.
▲
by
tzury
2y ago
In a recent interview , one of the founders claimed that one of Wiz smart moves was using a graph database for mapping cloud resources and their relations, while perhaps all other competitors used SQL or NoSQL. It helped them “get to the po
485.
▲
by
tzury
2y ago
Not the website's title. But indeed the common sense title. It is simply a tragic story, that represents many others, who simply happened and took place in smaller, "less famous" companies. Software is eating the world, and s
486.
▲
A VC Controlled Company Hijacked One of the Most Successful Open-Source Projects
(redis.io)
3 points
by
tzury
2y ago
|
2 comments
487.
▲
Unix: Making Computers Easier To Use – AT&T Archives film (1982) – Bell Labs [video]
(youtube.com)
2 points
by
tzury
2y ago
|
0 comments
488.
▲
by
tzury
2y ago
A. Within your circles. Make sure everyone knows you are available. B. Think of best creative solutions you came up with throughout your career. Use LLM to write a post every other day (!). Within a month, you have 15 posts, which people
489.
▲
by
tzury
2y ago
This is the earliest HN post available online: https://news.ycombinator.com/item?id=2
490.
▲
YC Founders Chat
(yc-founders-chat.vercel.app)
1 points
by
tzury
2y ago
|
0 comments
491.
▲
Detecting misbehavior in frontier reasoning models
(openai.com)
1 points
by
tzury
2y ago
|
0 comments
492.
▲
The MASK Benchmark: Disentangling honesty from accuracy in AI systems
(mask-benchmark.ai)
2 points
by
tzury
2y ago
|
0 comments
493.
▲
Structured data extraction from unstructured content using LLM schemas
(simonwillison.net)
2 points
by
tzury
2y ago
|
0 comments
494.
▲
by
tzury
2y ago
Copilot is a code completion, and authoring for developers. It’s not a presenting a copy of trained material, rather generating new content based on what it has learned in the context of the problem / query discussed in the prompt. Rep
495.
▲
by
tzury
2y ago
People who stayed at IBM because they could not afford going anywhere else. People who worked at companies acquired by IBM and could not afford going anywhere else. A mixture of both will be involved from now on in decision making regarding
496.
▲
by
tzury
2y ago
Thank you Steph! Your product is awesome, used continuously, and works well on all platforms (laptop, phone, etc). I am grateful and happy that I got the opportunity to thank you here.
497.
▲
Satya Nadells About the Majorana 1 chip breakthrough
(twitter.com)
1 points
by
tzury
2y ago
|
1 comments
498.
▲
by
tzury
2y ago
“Figures don’t lie, but liars can figure”. You can easily get drowned by a see of numbers and get confused and gaslighted, unless you don’t make sure all data is available and computable. Not sure how this release, which impressive by all m
499.
▲
by
tzury
2y ago
To my surprise, it is performing better than a ZMQ based solution I begun with (in terms of concurrent connections).
500.
▲
by
tzury
2y ago
I am currently experimenting with multi read/single write via async FastAPI endpoints and my benchmarks tells it will scale well enough for the first 6 month of anticipated growth, while major part of the time will be beta testing user
501.
▲
by
tzury
2y ago
Well, if you are long enough in the industry, in retrospective, most of the code, frameworks and platform you have worked on are now considered "legacy", given the dynamic fast pace of our environments. Data-centers -> Cloud M
502.
▲
Enigmaeval: A Benchmark of Long Multimodal Reasoning Challenges [pdf]
(static.scale.com)
1 points
by
tzury
2y ago
|
0 comments
503.
▲
by
tzury
2y ago
https://archive.is/6C2Tw
504.
▲
The Maze Is in the Mouse
(pravse.medium.com)
2 points
by
tzury
2y ago
|
1 comments
505.
▲
by
tzury
2y ago
Claude is indeed far more familiar amongst software engineers. Google Gemini integration into their docs/sheets/slides and Gmail perhaps will show different demographics in a few months, and that is yet before we heard from OpenAI
506.
▲
by
tzury
2y ago
Industry will get there pretty soon regardless of that bill or another, since there is a paradigm shift. The conversation is no longer about scraping bots versus genuine human visitors. Today’s reality involves legitimate users leveraging A
507.
▲
Ask HN: AI-Powered IDEs Streamlining Development–Or Making Us Work Harder?
5 points
by
tzury
2y ago
|
1 comments
508.
▲
by
tzury
2y ago
https://archive.is/etpZf
509.
▲
OpenAI Co-Founder John Schulman Leaves Rival Firm Anthropic
(bloomberg.com)
7 points
by
tzury
2y ago
|
3 comments
510.
▲
by
tzury
2y ago
The Demo: https://claude.ai/constitutional-classifiers
More ›