HN user

Zamaamiro

326 karma
Posts0
Comments61
View on HN
No posts found.

You probably didn’t write up a detailed prompt with perfect specifications in 10 seconds, either.

In my experience, it doesn’t matter how good or detailed the prompt is—after enough lines of code, the LLM starts making design decisions for you.

This is why I don’t accept LLM completions for anything that isn’t short enough to quickly verify that it is implemented exactly as I would have myself. Usually, that’s boilerplate code.

I’m working on a research cybersecurity tool that attempts to combine the natural language understanding and information synthesis strengths of LLM-driven agents with symbolic logic and knowledge bases expressed as Datalog programs for determinism and declarative semantics.

The approach is to perform system scanning using a combination of LLMs and traditional algorithms to dynamically populate a Datalog knowledge base. The facts of the program are constrained to a predefined “model schema” of sorts and a predefined set of rules that encode specialized domain knowledge of how new facts can be derived from known facts.

We generate proof trees / attack graphs from the knowledge base and queries posed to it. The attack graph uses big-step semantics to plan and guide the execution flow, and the system dispatches to agents with tool use to fill in the details and implement the small-step semantics, so to speak. This may include API calls to a Metasploit Framework server or RAG over vulnerability and exploit databases.

We use Pydantic AI to constrain the LLM output to predefined schemas at each step, with a dash of fuzzy string matching and processing to enforce canonicalization of, e.g., software names and other entities.

Tl;dr: neurosymbolic AI research tool for cybersecurity analysis and pentesting.

It's also worth noting that some of the damage will be permanent.

Trump's mini trade war with China in 2018 (for which he had to bail out farmers) led to US farmers permanently losing market share in soybean exports to Brazil.

In 2018, during Trump's first term, the U.S. and China engaged in tit-for-tat tariffs that led Beijing to take permanent steps to reduce its reliance on American farm goods.

The share of China's soybean imports from the United States dropped to 18% in the first 11 months of 2024, from 40% in the whole of 2016, while Brazil's share grew to 74% from 46%, according to Chinese customs data.

https://www.reuters.com/markets/commodities/chinese-buyers-s...

Yes. The port traffic data discussed in this thread is one such example.

I'm not an economist or supply chain expert myself, so I rely on actual subject-matter experts to interpret and contextualize the raw numbers for me. So far, they're all painting a pretty gloomy picture.

Good people to follow on this are CEO of Flexport Ryan Petersen [1], Jason Miller from MSU who had a great podcast with Derek Thompson [2], CEO of FreightWaves Craig Fuller [3], etc.

[1] https://www.businessinsider.com/small-businesses-buying-impo...

[2] https://www.theringer.com/podcasts/plain-english-with-derek-...

[3] https://www.cnbc.com/video/2025/04/28/its-about-to-get-much-...

That's not what trickle-down economics mean.

Trickle-down is the idea that giving tax breaks or benefits to the wealthy or big corporations will eventually benefit everyone else through increased investment or job creation. What I’m talking about here is the basic flow of goods and services in an economy—when that gets disrupted, the effects hit workers and consumers directly, not eventually or indirectly.

When imports slow down, businesses have fewer goods to sell or face higher input costs. That leads to higher prices for consumers and layoffs for workers. It doesn’t just impact “the rich.”

This isn’t about trickle-down economics; it’s about how supply chains work. The impact of these tariffs will show up in lost jobs, higher prices, and reduced access to everyday goods. Those are real effects for regular people, not just abstract economic concerns.

Reduced shipping volumes immediately mean less work for truckers—thousands of people taking a hit to their income right off the bat, which in turns leads to less economic activity. This is exactly what people mean when they say no part of the economy exists in a vacuum. It’s all connected.

It's an unpopular opinion because it's wrong.

No single aspect of the economy exists in a vacuum. Tariffs and lower shipping volumes portend thousands of small businesses potentially going under, translating to many more thousands of people losing their jobs and incomes. This has many unforeseen knock-on effects in declining economic activity.

This is a bad comparison.

Tobacco and alcohol, both of which have objective, measurable negative health outcomes supported by decades of research, versus some vague notion of "junk products" as defined by... who? And this is without even getting into the fact that the tariffs will raise the price of everything, not just these supposed "junk products."

Since when did free trade become a partisan issue?

All respected economists on all sides of the political spectrum agree that tariffs hurt the economy. The current President is the only political actor pushing for tariffs.

Reminding people that tariffs are a tax on the consumer is not a partisan issue; it is transparency and plainly economics 101.

This is quite a display of mental gymnastics and historical revisionism.

The last couple of years have seen a massive post-COVID backlash to incumbent parties all over the world—Democrats were just caught in the wave.

Why on Earth would anyone not call them anything but Trump Tariffs when he is the one who imposed them and when they’re uniquely his idea?