Hacker News: toplinesoftsys

New comment by toplinesoftsys in "Autoregressive Language Model on the 6502 Processor"

toplinesoftsys — Sun, 02 Aug 2026 23:27:27 +0000

This is amazing project! I hope it will result in real miniaturization of AI - for example, edge LLM inside of glasses. That will be awesome.

New comment by toplinesoftsys in "Let the machines in"

toplinesoftsys — Sun, 02 Aug 2026 22:11:02 +0000

I think one of key reasons why we have so many issues with LLMs is because there is no clear understanding of their limitations. This is how I see LLM limits:

1. AI is inconsistent and unpredictable. It can do what is expected today, but the same AI can do totally different, unpredictable, and maybe even dangerous thing tomorrow.

2. AI is bad on a long context - if you give AI instructions, it forgets/misses/misinterprets them. The longer the context, the more deviations are produced.

As a result, AI usage must follow these rules:

1. There should always be a human-in-the-loop to review every AI output - it solves problem #1. Yes, there will be a significant slowdown of the processes, but alternative is a quickly accumulating mess.

2. AI must be given only a short context to deal with - this is solution for problem #2. I predict this term will be popular soon - "AI microagent", meaning agent handling a short context. Yes, there will be a manual work integrating and orchestrating a huge number of these microagents, but it is still better than fixing all defects of AI-generated output after feeding it a single long context.

New comment by toplinesoftsys in "Karpathy’s Pelican"

toplinesoftsys — Sun, 02 Aug 2026 21:06:08 +0000

This is just a very rough basic game skeleton. Companies believing AI is smart enough to spit out almost ready product will learn a hard way that it will will give them just a starting point that still needs a lot of work. This is a nature of all modern LLMs - they easily lose context: the bigger the context the more losses and distortions are, especially in the middle. So, giving AI the complete book does not mean it will follow everything in it - quite the contrary.

New comment by toplinesoftsys in "As Reddit stock falls, CEO questions value of Google's AI Overviews"

toplinesoftsys — Sun, 02 Aug 2026 19:42:26 +0000

I am not sure there is a good solution for filtering out a synthetic content. AI detectors became basically useless, giving false positives a lot. Even human verification does not help - nobody prevents verified account to start using bots.

New comment by toplinesoftsys in "The first transatlantic telegraph cable was a bold, beautiful failure (2019)"

toplinesoftsys — Sun, 02 Aug 2026 19:38:59 +0000

This is a very familiar pattern: brilliant visionary bring new technology too early and it fails. The next guy does not need to be such a prodigy - he just need to be right on time to succeed.

New comment by toplinesoftsys in "Show HN: Fuse – statically typed functional programming language"

toplinesoftsys — Sun, 02 Aug 2026 18:12:18 +0000

Great project - clean and simple syntax, pure functional language, uses GRIN framework.

New comment by toplinesoftsys in "Ask HN: I still don't understand why AI agents need "skills""

toplinesoftsys — Sun, 02 Aug 2026 06:50:46 +0000

Skill is a document that AI agent ignores, distorts and forgets immediately after reading it :)

New comment by toplinesoftsys in "China's tech advances are causing chaos from Silicon Valley to the White House"

toplinesoftsys — Sun, 02 Aug 2026 06:46:40 +0000

Banning competition is the fastest way to lose the race.

New comment by toplinesoftsys in "Software for One"

toplinesoftsys — Sun, 02 Aug 2026 06:40:46 +0000

I love this approach. Simple ideas, specific problems, clean interface, right on point, took a week or one weekend to build. This is what software development industry was missing for sooooo long.

New comment by toplinesoftsys in "AI's real threat to jobs isn't job loss, it's lower paychecks, new research says"

toplinesoftsys — Sun, 02 Aug 2026 06:14:25 +0000

At the end of the day, the real threat of AI is degradation of businesses, eventually leading to extinction of those of them still absolutely convinced AI will replace everybody. The real AI adoption is extremely low, but many business leaders are (willfully?) blind. Eventually, the pendulum will swing in other direction and there will be enormous demand for human labour. It will happen when businesses finally start to realize that AI is ruining their businesses.

New comment by toplinesoftsys in "The Greenhouse and the Lens: Two Modes of Agentic AI Work"

toplinesoftsys — Sun, 02 Aug 2026 06:06:26 +0000

There is a third mode that requires combination of greenhouse and lens modes, and I did not see AI that is every remotely good working at that third mode. Unfortunately, many real applications are actually require the third mode, and I suspect this is one of key reasons why adoption of AI is so low.

New comment by toplinesoftsys in "Show HN: Elevators"

toplinesoftsys — Sun, 02 Aug 2026 05:58:11 +0000

This is a great food for thought next time I am waiting for elevator :)

New comment by toplinesoftsys in "China begins producing advanced chipmaking deep-ultraviolet lithography machines"

toplinesoftsys — Sun, 02 Aug 2026 05:52:17 +0000

I would not be overestimating this achievement, but not underestimating either. The truth is probably somewhere in the middle. China had access to the latest EUV tech for some time already, so they had abilities to reverse-engineer that tech. Eventually, the verdict will be clear only after we see actual chips produced using these machine, and when we learn their real yield.

New comment by toplinesoftsys in "The Art of 64-bit Assembly"

toplinesoftsys — Sun, 02 Aug 2026 05:49:10 +0000

We need books like that now more than ever. AI makes us lazy, forgetting how it works "under the hood". We should be curious and keep asking questions, or we will not be able to understand all this geneated AI code slope.

New comment by toplinesoftsys in "Linux desktop market share has hit over 10% in North America"

toplinesoftsys — Sun, 02 Aug 2026 05:45:24 +0000

Yes, the dam has been broken. Major linux distibutions are getting polished for many decades already and they are totally ready for the mainstream adoption. No amount of marketing will be able to reverse this trend. Considering that Big Tech is spending all their earned money (and even more by borrowing from investors) on AI, nobody will be even trying.

New comment by toplinesoftsys in "Show HN: I'm a 15 Year Old Wannabe Engineer, This Is a Cycloidal Gearbox I Built"

toplinesoftsys — Sun, 02 Aug 2026 05:31:08 +0000

Great project! Keep pushing, do not pay attention to anyone saying that AI will replace all software engineers.

New comment by toplinesoftsys in "The Kimi K3 Moment"

toplinesoftsys — Sun, 19 Jul 2026 00:59:38 +0000

One thing is absolutely clear - open-source models have already reached the level of top commercial ones. The last bottleneck is hardware, but that threshold is decrasing fast too. While it is still extremely expensive to run Kimi K3 model at home, there are already many very capable free models you can run on decent hardware. This trend will definitely continue.

New comment by toplinesoftsys in "Show HN: Powerful Visual Programming Language (Book)"

toplinesoftsys — Sat, 04 Oct 2025 03:46:27 +0000

For us, the goal is getting early feedback and potentially get into discussions to see where the idea stands. Benefits for others depend on their specific areas of interest. For example, people who work in areas of visual programming, low-code and AI code generation will be interested to know what new ideas impacting their domain come to light.

There is nothing preventing release of an idea before finalizing implementation. We are not the first going this route and I am sure there will be more after us. Like everything else, it has its cons and pros.

New comment by toplinesoftsys in "Show HN: Powerful Visual Programming Language (Book)"

toplinesoftsys — Sat, 04 Oct 2025 03:24:50 +0000

The product here is the book, not software (yet). You can download and read it.

New comment by toplinesoftsys in "Show HN: Powerful Visual Programming Language (Book)"

toplinesoftsys — Sat, 04 Oct 2025 01:21:51 +0000

The core principle of Pipe is that it is not trying to replace text-based languages. There will always be areas where text is better. What Pipe is doing is visualizing structures on top of textual languages. It means that mouse manipulations in Pipe are going to be significantly less intensive because Pipe does not try to implement x=x+1 visually (I think it is the key mistake in majority of existing VPLs including LabVIEW and hence physical fatigue). What Pipe does is encapsulating non-visual code that does x=x+1 into a reusable visual element.

Also, AI can generate textual code calculating x=x+1. It means we mostly do not even need to touch textual code - AI will do it for us perfectly. All we need to do is converting AI-generated code into visual components for composing visual workflows.

To sum up, AI and visual programming create a perfect synergy: AI generates code and visual programming composes generated fragments into visual components. This should minimize physical efforts needed to manipulate visual workflows as devs are going to work only on high (visual) level of abstractions while AI will be providing low (text-based) abstractions.