westoque

Comparing agentic Qwen3.6 35b to Claude Opus is like a junior with knowledge across the board, that you really need to guide, versus a senior that thinks with you on architecture.

that's why i use the frontier models because its a senior co-worker vs a junior. if you use the junior for the sake of privacy i think you're missing out on the best insights for a specific task.

There's a formula I've bookmarked from someone to help you get better social capital. In order to increase your surface for social capital you need to expand your influence, therefore you need to increase your luck:

L (luck) = D (doing) x T (telling)

Basically, you need to be able to show you are able to do stuff (D) and with enough people seeing this by you sharing (T) through different channels, can open doors. I haven't been doing it as heavily but it has already given me incredible opportunities.

i've been thinking about this as well. but having built a startup, i've learned that users don't care as long as they are given the value and most convenience. they don't really care much at security as much as we do. just look at openclaw? but maybe it's our job to make sure it is taken care of vs assuming the user cares and just make it look seamless.

my takeaway from this is that it should now be MANDATORY to have an LLM do a scan on the entire codebase prior to release or artifact creation. do NOT use third party plugins for this. it's so easy to create your own github action to digest the whole codebase and inspect third party code. it costs tokens yes but it's also cached and should be negligible spend for the security it brings.

The problem for me is that the development practices of the people that are working on it are suboptimal at best; they're constantly releasing at an extremely high cadence, where they don't even spend the time to test or fix things (or even build a proper list of changes for each release), and they add, remove, refine, change, fix, and break features constantly at that accelerated pace.

this is what i notice with openclaw as well. there have been releases where they break production features. unfortunately this is what happens when code becomes a commidity, everyone thinks that shipping fast is the moat but at the expense of suboptimality since they know a fix can be implemented quickly on the next release.

one thing it did massively for me was save me time from questions like should i go with X or Y option questions. before i used to just think longer about tradeoffs but with AI it became a lot faster. no more procrastination due to decision fatigue.

I simply booted up a VM with an H100, ssh’d into it with Cursor, and prompted the agent to set up an inference server that I could ping from my web generation app. What used to take hours or days of painful, slow debugging now takes literally minutes.

an awesome takeaway from this is that self-hosted models are the future! can't wait for hardware to catch up and we can do much more experiments on our laptops!

i really think this is part of the pitch deck for bun's funding. that a bigger company would acquire it for the technology. the only reason an AI company or any company for that matter would acquire it would be to:

1. acquire talent.

2. control the future roadmap of bun.

i think it's really 1.

Current AI image models can't understand grid-based pixel art.

sounds like a good use case to fix this problem from the model layer. an image gen model that is trained to make pixel perfect art.

i like how claude code currently does it. it asks permission for every command to be ran before doing so. now having a local model with this behavior will certainly mitigate this behavior. imagine before the AI hits the webhook.site it asks you

AI will visit site webhook.site..... allow this command? 1. Yes 2. No

a concern i have is that it's only a matter of time before a similar attack is done to electron based apps (which also have packages installed using npm). probably worse because it's installed in your computer and can potentially get any information especially given admin privileges.

Recently launched an AI cafe ordering app https://www.hellocafe.ai using chat and voice.

We wanted to make concept for an app using all local models for chat (llama 3.1 8B) and voice (whisper). Deployed using kubernetes and easily scalable not to mention fully open source!

Refreshing that it's built using the vue framework vs the typical react.

They're getting my $200 a month and it's easily the best money I spend

Can you explain? I don't see how $200 makes that much difference than what I get from paying $20/month with OpenAI? What's the use case?

Came here to say this. What makes a great general purpose language is not only in the programming part but also in transportability. I think Go takes the cake here. If only Elixir apps could be compiled and transported as a binary to similar systems and it works, then it would make it a great general purpose language. Until then, it's really only a good client/server language as it was intended to be.

Such a great quote. Mostly true if viewed especially from a business standpoint. I for one also see code as creative expression, a form of art. I like coding because I can express a solution in a way that is elegant and nice to read for myself and others. A bit shallow but If you've read code that is written elegantly, you'll know that immediately.

Must be my ignorance but everytime I see explainers for LLMs similar to the post, it’s hard to believe that AGI is upon us. It just doesn’t feel that “intelligent” but again might just be my ignorance.

This weekend I’m working on mobile app that allows you to upload photos and it turns everything into a stitched anime (ghibli or not) with movements and eventually sound and script. Great to make mini animes from your day or travels or anything.

I vibe coded the above including everything: code, design, logos. Just did it solo. It has all error handling, video generation notifications (it takes a while) and credit system. I myself can't believe it's been done in a month with AI. It's already in closed beta in iOS and android app stores. Let me know if you want to try it out before public release.

My quoted comment above was 28 days ago. This is working on this part time and with a family.

EDIT: Added context.

This weekend I’m working on mobile app that allows you to upload photos and it turns everything into a stitched anime (ghibli or not) with movements and eventually sound and script. Great to make mini animes from your day or travels or anything.

not to hijack this thread but my dad did extensive research in sea breeze and rainfall modeling and he would have loved to see these AI and machine learning advancements in weather prediction.

[0] https://www.revistascca.unam.mx/atm/index.php/atm/article/do...

[1] https://wmo.int/about-wmo/awards/international-meteorologica...

working on cursor for desktop. why rely on AI agent that’s self-contained when it’s limited, can’t access the browser, can’t open apps or click around.

i simply want mine to be able to fill in forms in preview with a passport image as context. also to be able to do recurring tasks as if i was the desktop user. e.g., i’m going to bed keep working on this spreadsheet.

it’s working and built but very slow and buggy atm. uses multimodal LLMS and OCR but lots more optimizations needed. need to make it a lot faster. can demo it and need help if anyone is interested.

i've worked on many legacy code bases and i think an underrated skill that you need is courage.

courage to be able to push changes you think could be breaking changes but always be optimistic that you to be able to fix as soon as possible. tests are ideal, but not everyone has 100% coverage. sometimes you just have to trust your gut and do diligence and push the change. don't be scared to push something because you don't know the behavior in production. the only way to know is to try.

reminds me of the anthropic claude jailbreak challenge which only pays around $10,000. if you drive the price up, i'm pretty sure you'll get some takers. incentives are not aligned.

I'm a CTO and scaled multiple companies from zero to multiples of TBs of data (also to millions in revenue). I also worked on companies such as yourself that have an existing solution and want to revamp their entire technology base.

The short answer is yes, you should hire an in-house team so you can execute faster and also reduce the red tape of the provider not being more proactive.

How to do this is more nuanced:

1. I would start by documenting and/or asking for documentation on how the whole system works. This is key. I would start with a high level diagram. Even if you're not technical, this would allow you to evaluate the existing stack and see what could be worked on in parallel to the existing tech if any. I feel like you are technical enough to understand this since you understand CRUD and other terminologies. You can even set a time for the provider to explain the services to you if needed. I say parallel because I'm assuming you can't just quit using the system entirely, it will need to function until you do the switchover.

2. Once you have a good understanding of the system, this will allow you to be more specific on your needs. I almost never recommend a re-write but based on the statement that you are using a third-party product, this might be the case here. If it is indeed a re-write, I would approach it so that it's 1-1 parity with the existing system first, and if not, maybe with some ample planning and product grooming make the features even better. Furthermore, this will allow you to make a basic PRD for the requirements of what you are building. Doesn't have to be really specific but it will make it clear to yourself and others on what the task would be.

3. Hiring. At this point, I would start hiring a CTO/VP/senior engineer. Describe the problem from 2., maybe deep dive even if needed and start to get technical leadership going. I'm assuming cash is not a problem so if you start at $175k/year salary (maybe some stock/benefits) you can find many technical leaders that would go for this. Once you make that first hire, if they have a good network and or social clout it should be easy for them to hire people and/or consultants to do the job.

4. Iterate. Once you have that amazing team and if they have a good process in place they should keep iterating on the product and continue doing this on a steady pace.

5. Once the product is ready or any point that is it deployable make the switchover. At this point, you should have full in-house control of the product and have competent tech team in hand to do whatever is needed.

NVDA market cap for sure is bonkers but AI is here to stay. Anytime you question the future, think if you can see yourself in "that" future. For me, AI is already a part of my workflow and it's beneficial to my daily life. From coding, explainers, and to random questions in perplexity.

In coding, it made me a lot more productive. No more do I have to scour Google or Stack Overflow. Instead, I get the answer straight from the AI. It might not be working at times but it's definitely better.

Even random questions like "What is the cost to produce the Apple Pro Display XDR?" is being answered more promptly by AI.

This is contrary to the crypto craze where there was expected use but sadly it did not come to fruition. And until now, it's still a speculative industry, a solution looking for a problem.

just basing my assumptions on:

- https://www.theregister.com/2024/02/08/apple_web_apps_eu/

- https://itnext.io/apple-wants-to-kill-pwas-0895be2e497b

i really hope they don't as it does provide an acceptable alternative to native apps. not everyone has a mobile team that can build native. imagine a small team just making a web app could now also provide a similar experience and value for mobile out of the box.

alternatively you can get away from the app size bloat by using X as a PWA (progressive web app). in iOS, this would use safari as the app container and you would just have the assets to download for dependencies. it's sad however that apple is now trying to kill PWAs. i wish people would use PWAs more as this actually makes it easier to build apps.

Most of the speed I believe is from using the server iodine https://github.com/boazsegev/iodine which is a wrapper around facil.io https://facil.io that is built using C.

Great project either way!

I was using the Kinesis Advantage 2 (KA2) for the longest while 8+ years now, and then recently upgraded to the Kinesis Advantage 360 and the KA2 was still hands down the best keyboard that I used. I still switch between the KA2 and KA360 but the KA360 is as mentioned plagued with so many issues, from the bluetooth connection issues all the way to setting the keymap using the ZMK software. I would still recommend the KA2 for most users and also the best bang for the buck if you can get one used or refurbished. The split also doesn't make as big of a difference as it seemed at least from my proportions.

regarding concurrency, language plays an important role and pretty much dictates how code is written which is i believe where it has most of the frustration. in erlang it’s in a functional style, in javascript it’s in an asynchronous style. what i’ve come to realize is that it’s still better and more maintainable to think synchronously and have the core of the tech handle concurrency, for example golang with it’s goroutines or multi process in ruby/python. admittedly it’s not as concurrent/distributed as erlang but should be easier to work with.