akrylov

lpf is GitOps for Linux networking: PF-style firewall, NAT, QoS, and routing policy with live diff, guarded apply, and automatic rollback for remote changes.

I have not tried deerflow 2.0 but the community showcase is not particularly impressive to be honest. Single agent system can do just as good. Any multi-Agent framework must constantly outperform single-agent on a different tasks.

The name of the thread is provocative, but the premise is valid - I have yet to see anything produced by multi-agent frameworks (langchain or bespoke works) that produced value. Anthropic pushes vibeCAD, vibeVFX, vibePowerPoint but the results are underwhelming. The real value is in codegeneration and autonomous infra, research.

No US has been "winning" or rather leading thus far, but there is no guarantees that it will ever "win". I do not subscribe to idea of omnipotent omnipresent AGI. China plays a long game, I think DeepSeek does not engage in platform building on purpose. DeepSeek was probably assigned with a role of a primer lab, the goal is to replace CUDA, align with Huawei chips to do cutting edge research and cross-pollinate other teams in China. They might even hide the best models on purpose. In a long run China will use its industrial capability to apply and use AI better than anyone else. And that would be a good thing for the World.

It's not "AI bubble" - at this point it's a software bubble. It's Antropic or OpenAI that should justify their valuations, they have close to billion customers at this point. It's non-AI software companies without strong cloud business that must justify why their core product is not going to be replaced with the help of LLMs. It's not "fair" but that is how it's seen by the Wall Street.

>> And, yet, the US AI companies are not actually making a profit, right?

I think they already, actually making profits especially Antropic. But think how important it's from a business standpoint - the entire software stack from OS to Databases to browsers will be rewritten in the near future, for a company such as Oracle or IBM it means their bread and butter/cash cow can be replaced. It's worth almost any kind of Capex. And from Washington standpoint it's more important than F-35 program or even Apollo mission.

No, Alibaba is excellent top-5 easily.

The trade war and tariffs are bringing inflation, consumer prices will soar, but from a geoeconomical standpoint this will hurt China (And EU) more than US. US consumer on average has the deepest pockets in the World and people the top will make money on insider trading, stock anyways. If AI tokens will become like US dollar, it will be under total control of the Fed.

True, I would have preferred benevolent dictator scenario, like with the Internet. But this time around it's different - AI data centers will be like protected like embassies.

Jobs, Wozniak, Gates - it's a myth that you need poor migrants pulling themselves by their bootstraps to innovate. Sometimes a nazi scientist like Wernher von Braun is what it takes.

It's a myth. IBM, Xerox, HP, DEC was innovating long before H1B's.

The goal is global domination as always, unfortunately. DARPA and the Pentagon helped create the Internet, and Silicon Valley later turned it into a major commercial success.

In a capitalist society it's all about bottom line. OpenAI and Antropic, Amazon, Nvidia winning bit time. And I would have preferred it to be only about money, but US deep state certainly will capitalize on it like on US dollar (sanctions) and cloud infra (Cloud Act).

Terraform and Ansible are powerful tools for infrastructure automation, but they fall short in handling complex, long-running workflows, state management, and fault tolerance. Modern DevOps tooling often struggles with scalability, error recovery, and dynamic orchestration across distributed systems. A workflow engine addresses these gaps by providing robust orchestration, retries, checkpointing, and multi-tool integration, making it a prerequisite for scaling operations in dynamic, cloud-native environments.

Modern Kubernetes deployment methodologies have grown increasingly complex, layering abstraction upon abstraction in pursuit of flexibility. This article challenges that trajectory by examining how fundamental Unix tools combined with Makefiles can provide a more transparent and maintainable alternative to popular solutions like Helm and Kustomize.

Most blog posts focus on using top-tier LLMs or setting up complex AI pipelines for large corporations. But what if your data is private, and you don’t have access to top-tier ML talent or massive infrastructure? In this article, we show how to fine-tune a model for mid-sized software development teams or IT support, using your own domain expertise. With Apache Answer and InstructLab, you can build a powerful, cost-effective AI solution tailored to your specific needs.

Most secret-scanning tools for Git repos have a big problem: their JSON output is often broken or hard to parse, making automation a pain. They find secrets just fine, but trying to use their output in a CI pipeline or backend service usually means wrestling with malformed data or writing custom parsers.

That’s why I built a lightweight tool in Julia that does two things right: it outputs clean, valid JSON, and it lets you define custom detection patterns via a YAML config file. No more hacking around messy outputs or hardcoded rules—just structured results that actually work in automated workflows.