soltanov

5 ms·

61.
Measuring Malicious Intermediary Attacks on the LLM Supply Chain(twitter.com)
1 points by soltanov 1mo ago|0 comments
62.
DeepSeek v4.1 Flash Uncensored(huggingface.co)
60 points by soltanov 1mo ago|28 comments
63.
A New Chess Benchmark for Language Models(chessbench.ai)
2 points by soltanov 1mo ago|0 comments
64.
Gemini 3.8 Flash passed Claude Fable 5.1 on this bench(chessbench.ai)
2 points by soltanov 1mo ago|0 comments
65.
by soltanov 1mo ago
fully agree. cant even started to read this. just skipped directly.
66.
If PHP Were British(aloneonahill.com)
1 points by soltanov 1mo ago|0 comments
67.
by soltanov 1mo ago
Google treats its customers so blindly. in short, Google is not user centric company at all.
68.
by soltanov 1mo ago
First they need fix their ai based tools. Paid users of Google AI are exhausted from Google.
69.
Mapped: AI Adoption by Country in 2026(visualcapitalist.com)
2 points by soltanov 1mo ago|0 comments
70.
by soltanov 1mo ago
Task success is not enough; reliability and cost matter.
71.
by soltanov 1mo ago
Agent systems become most credible when they produce artifacts that can be independently checked, not when they merely produce persuasive explanations.
72.
by soltanov 1mo ago
This is not only a fairness problem. It is an agent-memory problem: a system can mistake its own early choices for evidence.
73.
by soltanov 1mo ago
When a web page, email, or document is marked untrusted, does that label persist through summaries and tool calls?
74.
The SaaS Subscription Trap: How "Forgot to Cancel" Became a Business Model(indiehackers.com)
2 points by soltanov 1mo ago|0 comments
75.
Fx2-Cmix-Transformer(mattmahoney.net)
1 points by soltanov 1mo ago|0 comments
76.
by soltanov 1mo ago
It is useful architecture tool.
77.
by soltanov 1mo ago
"Reply on HN" thing doesnt work?
78.
New off-peak rates: 50% lower prices for DeepSeek-V4(ollama.com)
4 points by soltanov 1mo ago|0 comments
79.
by soltanov 1mo ago
I cant event tell if you are being sarcastic or not :)
80.
Ollama's Transparent Pricing(ollama.com)
3 points by soltanov 1mo ago|0 comments
81.
The Perpetual Tourist Experiment(experimentalliving.substack.com)
1 points by soltanov 1mo ago|0 comments
82.
AI Can Help Plan a Bioweapon(secondthoughts.ai)
3 points by soltanov 1mo ago|0 comments
83.
by soltanov 1mo ago
Standart Google Tradition
84.
Ollama Plans Changed(ollama.com)
3 points by soltanov 1mo ago|0 comments
85.
by soltanov 1mo ago
Title can be "Codes - typed application return values for Kotlin and Java".
86.
by soltanov 1mo ago
is google fixed usage quota in antigravity? or still is this a problem of antigravity?
87.
by soltanov 2mo ago
Bill Gate's post on HN: https://news.ycombinator.com/item?id=49447057
88.
The World Needs a Plan(garymarcus.substack.com)
4 points by soltanov 2mo ago|1 comments
89.
by soltanov 2mo ago
Long readme and formatting is hard to read. but project is interesting. will watch on this
90.
by soltanov 2mo ago
@CliffStoll did you talk with Markus Hess. i think he is active on thewilyhacker.de?

More ›