msukhareva

6 ms·

1.
by msukhareva 28d ago
Explaining in plain language about harnesses, logs, models and agentic loops. No civilisations, no hive mind, no fearmongering
2.
by msukhareva 6mo ago
The most “moral” of all moral companies, Anthropic, may not be able to stop building capabilities that look very much like weapons. This time it is cyber weapons. In the latest AI Realist article, I go through the Mythos system card. 245 pa
3.
by msukhareva 7mo ago
Making the Pentagon dream of maximising lethality true
4.
Free Iran War Monitor(iran.airealist.org)
1 points by msukhareva 7mo ago|0 comments
5.
Iran and Region Monitor of Attacks and Major Events(newsfeed-staging.pages.dev)
2 points by msukhareva 7mo ago|0 comments
6.
by msukhareva 8mo ago
Daniela was just about to begin her studies in finance. Before the semester started, she went to a music festival with her boyfriend, Daniel. Daniela and Daniel were among 387 people killed by Hamas in October 2023. Today, her mother finds
7.
by msukhareva 8mo ago
Europe has started preparing for a US tech shutdown. What seemed impossible yesterday is becoming increasingly likely. European dependency on US tech is a leverage that Trump knows how to use. But what happens if he decides to cut it off?
8.
by msukhareva 9mo ago
The latest openAI image model when asked to generate a list of animals creates cursed pokemons. The reason for this is the new state-of-the-art architecture of their image generation model. The AI realist article "ChatGPT creates monst
9.
by msukhareva 9mo ago
I built an architecture that enables continuous learning for open-weight models. The "curse" of LLMs is the knowledge cut-off. RAG is just a band-aid. Google's "Titans" memory solves this, but costs millions to trai
10.
by msukhareva 10mo ago
Gartner just published a guide to burning AI budget. It is called “Top 10 Strategic Technology Trends for 2026.” In the next weeks it will land on every CIO and CTO desk, get copied by every consulting PPTX factory, and then get digested by
11.
by msukhareva 11mo ago
it is time to call it. Their open weight models are the ones that keep research alive and enable startups and even bigger companies to build AI solutions while keeping control of the model. The Western companies chose the path of hype and p
12.
by msukhareva 11mo ago
Nvidia stock goes up because of Bitcoin. How many times have you heard this phrase? And yet, ironically, crypto was never the main driver of Nvidia’s success. At some point, mining Bitcoin on GPUs wasn’t the best method anymore. The role of
13.
by msukhareva 1y ago
haha, that is funny. I did not think about it but I do not think it will work every time
14.
Technical Explanations Why LLMs Use Em Dashes(msukhareva.substack.com)
2 points by msukhareva 1y ago|0 comments
15.
The E-Commerce Trap: How ChatGPT Tracks You(msukhareva.substack.com)
4 points by msukhareva 1y ago|3 comments
16.
by msukhareva 1y ago
OpenAI urgently needs to make money. Its most valuable product isn’t the LLM - it’s the users. The company has now set up an efficient way to track your behavior online and sell that data to marketers. The setup might look harmless - a tiny
17.
by msukhareva 1y ago
I am not even surprised. Eventually all this company will stop talking about AGI and curing cancer and will just turn LLMs into another way to sell stuff to people, just like OpenAI already did
18.
by msukhareva 1y ago
RAG was always somewhat of a Frankenstein combining two things that should not be combined: information retrieval based on string matching enhanced with embeddings and LLM that needs not string matching but informative texts. If string matc