Andrew Savetchuk

Software Engineer

[Kalshi @Kalshi]

JUST IN: Intel to launch new AI chip to rival Nvidia and AMD

[Bharat Kara @KaraBharat]

The perfect landing page setup

[unusual_whales @unusual_whales]

Jeff Bezos: Smart people about [AI saying job will be replaced] are dead wrong

[LM @loyalmoses]

Just weeks after GitHub announced they were going to have AI scanning their systems.

[Andrew Savetchuk @AndrewSavetchuk]

Microsoft installed Teams on my personal PC through a Windows update. Why are new applications being installed through Windows Update without prompting the user first?

[unusual_whales @unusual_whales]

BREAKING: Apple, $AAPL, and Intel, $INTC, announced a preliminary deal for Intel to manufacture select chips used in Apple devices, per WSJ

[Emil Privér @emil_priver]

Yes, we're aware

[Gergely Orosz @GergelyOrosz]

Pull requests disappeared on GitHub for many (all?) users.

This is just the latest outage on a platform where reliability has been beyond unacceptable the last few months.

A fair question: at what point would customers move? How much pain is too much? And where do they move?

[Polymarket @Polymarket]

NEW: Apple's Weather app is experiencing a major outage.

[Andrew Savetchuk @AndrewSavetchuk]

Great talk on software fundamentals and AI by @mattpocockuk

https://t.co/ddyhA16T1d

[unusual_whales @unusual_whales]

AI chatbots misdiagnose in over 80% of early medical cases, per FT.

[Andrew Savetchuk @AndrewSavetchuk]

Claude Code is back on the Pro plan

[Wes Bos @wesbos]

Looks like the $20/mo plan no longer include Claude Code. Smallest plan is now $100/mo

[Andrew Savetchuk @AndrewSavetchuk]

> scored 100%
> solved 0 tasks

Leaderboard comparisons may be noise. Don’t trust the number. If a system can score ~100% without solving the problem, the benchmark is measuring the wrong thing.

https://t.co/gN6P0jbPcW

[Hao Wang @MogicianTony]

SWE-bench Verified and Terminal-Bench—two of the most cited AI benchmarks—can be reward-hacked with simple exploits.

Our agent scored 100% on both. It solved 0 tasks.

Evaluate the benchmark before it evaluates your agent. If you’re picking models by leaderboard score alone, you’re optimizing for the wrong thing. 🧵

[Douglas A. Boneparth @dougboneparth]

My daughter’s lemonade stand was struggling, so we pivoted to AI compute.

She’s been acquired by Microsoft for $1.2 billion.

[Andrej Karpathy @karpathy]

LLM Knowledge Bases

Something I'm finding very useful recently: using LLMs to build personal knowledge bases for various topics of research interest. In this way, a large fraction of my recent token throughput is going less into manipulating code, and more into manipulating knowledge (stored as markdown and images). The latest LLMs are quite good at it. So:

Data ingest:
I index source documents (articles, papers, repos, datasets, images, etc.) into a raw/ directory, then I use an LLM to incrementally "compile" a wiki, which is just a collection of .md files in a directory structure. The wiki includes summaries of all the data in raw/, backlinks, and then it categorizes data into concepts, writes articles for them, and links them all. To convert web articles into .md files I like to use the Obsidian Web Clipper extension, and then I also use a hotkey to download all the related images to local so that my LLM can easily reference them.

IDE:
I use Obsidian as the IDE "frontend" where I can view the raw data, the the compiled wiki, and the derived visualizations. Important to note that the LLM writes and maintains all of the data of the wiki, I rarely touch it directly. I've played with a few Obsidian plugins to render and view data in other ways (e.g. Marp for slides).

Q&A:
Where things get interesting is that once your wiki is big enough (e.g. mine on some recent research is ~100 articles and ~400K words), you can ask your LLM agent all kinds of complex questions against the wiki, and it will go off, research the answers, etc. I thought I had to reach for fancy RAG, but the LLM has been pretty good about auto-maintaining index files and brief summaries of all the documents and it reads all the important related data fairly easily at this ~small scale.

Output:
Instead of getting answers in text/terminal, I like to have it render markdown files for me, or slide shows (Marp format), or matplotlib images, all of which I then view again in Obsidian. You can imagine many other visual output formats depending on the query. Often, I end up "filing" the outputs back into the wiki to enhance it for further queries. So my own explorations and queries always "add up" in the knowledge base.

Linting:
I've run some LLM "health checks" over the wiki to e.g. find inconsistent data, impute missing data (with web searchers), find interesting connections for new article candidates, etc., to incrementally clean up the wiki and enhance its overall data integrity. The LLMs are quite good at suggesting further questions to ask and look into.

Extra tools:
I find myself developing additional tools to process the data, e.g. I vibe coded a small and naive search engine over the wiki, which I both use directly (in a web ui), but more often I want to hand it off to an LLM via CLI as a tool for larger queries.

Further explorations:
As the repo grows, the natural desire is to also think about synthetic data generation + finetuning to have your LLM "know" the data in its weights instead of just context windows.

TLDR: raw data from a given number of sources is collected, then compiled by an LLM into a .md wiki, then operated on by various CLIs by the LLM to do Q&A and to incrementally enhance the wiki, and all of it viewable in Obsidian. You rarely ever write or edit the wiki manually, it's the domain of the LLM. I think there is room here for an incredible new product instead of a hacky collection of scripts.

[unusual_whales @unusual_whales]

NEWS: Anthropic is preparing for the release of its new Opus 4.7 model, per the Information

[Andrew Savetchuk @AndrewSavetchuk]

Interesting read about LLMs and "Trendslop".

"Leaders might assume that LLMs are able to offer a kind of unbiased, outside perspective. But new research found that leading LLMs have clear biases when it comes to strategy and consistently recommend strategies that align with modern managerial buzzwords and trends rather than context-specific strategic logic."

"Across thousands of simulations, we saw LLMs almost uniformly select the same trendy strategies, regardless of context. We call the propensity for AI to opt for buzzy ideas over reasoned solutions “trendslop”."

https://t.co/ZInk1rjiI7

[Andrew Savetchuk @AndrewSavetchuk]

Mythos could be a sales pitch ahead of the IPO. I’d take it with a grain of salt.