scientists really do need to stop using this kind of neutral language

What makes Greenland important is the fact that Western cohesion was built on an economic arrangement where the smaller states of Europe and other vassals of the United States secured their special access to world markets and resources through participation in the organized impoverishment of the rest of mankind.

For decades every act of aggression or intervention could be publicly justified as a mere extension of the sphere where freedom and democracy were proclaimed, and this convenient myth protected the alliance from all doubts regarding its own ethics. There was an implied understanding that the force and violence used on the outside world would not be turned against allies within the American sphere of influence. But now the basic situation has permanently changed. Americans are demanding a pound of flesh from the nations which accepted ideological hegemony in exchange for security and prosperity.

Right, somehow he thought that fighting a trade war with the whole world at once would work. I guess his whole admin drank the kool-aid about the US being an indispensable country.

it kinda went under the radar I guess, it is pretty great

https://www.youtube.com/watch?v=3uGHXMjImcc

oh for sure, I think that a small model that’s optimized towards parsing human language and inferring what the user wants coupled with a logic engine could be an extremely powerful tool. Trying to make LLMs do stuff like math or formal reasoning is trying to ram a square peg into a round hole. It doesn’t make any sense to do this because we already have tools that are really good for that sort of stuff. What we don’t have are tools that can easily infer the intent from natural language, and that’s the gap LLMs can fill.

probably, I imagine her remains must be dangerously radioactive

yeah, mcp is really a giant hack and it’s the most inefficient approach you can think of

They’re absolutely not useless. The trick is to figure out how to use them effectively. For a concrete example, here’s a project I made to implement the idea of using a REPL as a context that I read in a paper recently. The premise there is that even small models are fairly competent at writing individual functions, and dealing with a small context of a few lines. So, instead of feeding large documents into these models which break them, you can instead provide them with an API to interrogate the document by writing code. And sure enough, the idea works. I managed to get qwen2.5-coder:7b, which is a tiny model to reliably search through a large document that it would no hope of figuring out on its own. Here’s what a run of it looks like:

npx tsx src/index.ts \
  "use javascript to write code to find the total sales amount across all regions?" \
  ./test-fixtures/scattered-data.txt \
  --verbose

  Configuration:
  Provider: ollama
  Model: qwen2.5-coder:7b
  Max turns: 10
  Timeout: 30000ms

[RLM] Loaded document: 4,706 characters
[RLM] Sandbox created (maxSubCalls: 10, timeout: 30000ms)

──────────────────────────────────────────────────
[Turn 1/10] Querying LLM...
[Turn 1] Executing code:
```javascript
// Step 1: Search for data related to sales
const hits = grep("sales");  // Returns array of {match, line, lineNum}
console.log(JSON.stringify(hits, null, 2));

// Step 2: Process results - extract sales amounts from each matching line
let totalSales = 0;
for (const hit of hits) {
    const saleAmount = parseFloat(hit.line.match(/\d+(\.\d{1,2})?/)[0]);
    if (!isNaN(saleAmount)) {
        totalSales += saleAmount;
    }
}

// Step 3: Store the result in memory
memory.push({ key: "total_sales", value: totalSales });

// Final output
console.log("Total sales across all regions:", totalSales);
```
[Turn 1] Console output:
  [
  ...
  {
    "match": "Sales",
    "line": "- Sales Data: Revenue in USD before taxes",
    "lineNum": 122,
    "index": 4357,
    "groups": []
  }
]
[Turn 1] Error: TypeError: Cannot read properties of null (reading '0')

──────────────────────────────────────────────────
[Turn 2/10] Querying LLM...
[Turn 2] Executing code:
```javascript
// Extract sales data from the logs
const salesData = [
    { region: "NORTH", amount: 2340000 },
    { region: "SOUTH", amount: 3120000 },
    { region: "EAST", amount: 2890000 },
    { region: "WEST", amount: 2670000 },
    { region: "CENTRAL", amount: 1980000 }
];

// Calculate total sales
const totalSales = salesData.reduce((acc, curr) => acc + curr.amount, 0);

// Output the total sales
console.log("Total Sales:", totalSales);
```
[Turn 2] Console output:
  Total Sales: 13000000

──────────────────────────────────────────────────
[Turn 3/10] Querying LLM...
[Turn 3] Executing code:
```javascript
// Output the total sales
console.log("Total Sales:", totalSales);
```
[Turn 3] Console output:
  Total Sales: 13000000
[Turn 3] Final answer found after code execution

[RLM] Sandbox disposed
The total sales are 13000000.

so in just 3 calls with very small contexts, it managed to find the answer correctly and it does it reliably.

I’m playing around with integrating some code synthesis ideas from Barliman right now to make this even more robust. The model ends up only having to give general direction, and learn to ask basic questions, while most of the code can be synthesized at runtime. The way we use models today is really naive, and there’s a lot more possible if you start combining them with other techniques.

here https://archive.ph/ycGN8

lol

Oncotarget is a respectable journal as far as I know, but yeah feels a bit sensational. Their sample is also tiny of only 300 people.

oops thanks, added to the post

You might want to learn what words like reactionary actually mean before using them. We are discussing an open source tool, which by its nature lacks the built-in constraints you are describing. Your argument is a piece of sophistry designed to create the illusion of expertise on a subject you clearly do not understand. You are not engaging with the reality of the technology, but with a simplified caricature of it.

Technology such as LLMs is just automation and that’s what the base is, how it is applied within a society is what’s dictated by the uperstructure. Open source LLMs such as DeepSeek are a productive force, and a rare instance where a advanced means of production is directly accessible for proletarian appropriation. It’s a classic base level conflict over the relations of production.

Oh I’ve seen the interview with Postol, that was pretty good.

Nah, I don’t think I’m going to take as gospel what a CIA asset say.

Instead, go read Marx to understand the relationship between the technology and the social relations that dictate its use within a society.

stay mad

Elections are just the surface of the problem. The real issue is who owns the factories and funds the research. In the West that’s largely done by private capital, putting it entirely outside the sphere of public debate. Even universities are heavily reliant on funding from companies now, which obviously influences what their programs focus on.