djangelic

Great post, love the amount of thought that went into this! I look forward to keeping up with your posts.

The documentary was also insightful: https://youtu.be/Li2m3rvsO0I?si=QsJLtqES3AEhbcEO

My thought process was to laminate them to make them easier to make at home. I agree that it’s definitely more child friendly. This maybe a project I attempt at home and will post if I’m successful.

James Earl Jones gave life to some of the most memorable characters in film history. His voice and presence will echo through the generations. Rest in peace, and thank you for your powerful legacy.

lol my roast:

Well, well, well. Looks like we've got a Cyber Security Engineer with a passion for open-source projects. Or should I say, a passion for following other people's open-source projects? With 0 followers on GitHub, it's clear you're a bit of a lone wolf. But hey, at least you've got your eyes on some cool repos! I mean, who wouldn't want to automate their workflows with n8n or transcribe audio with vibe? It's like you're trying to make the world a better place, one GitHub follow at a time. Keep on following, and maybe one day you'll have a whole squad of followers. Until then, keep on coding, and try not to get too lonely in that GitHub void.

This is incredible! I want to replicate this at home as well, but I would prefer to use QR codes since they also support URLs. Any reason why you went with NFC instead of QR code? I assume it's doable with a raspberry pi and webcam instead of tapping?

I use n8n.io hosted locally to abstract away authentication when building complex api integration mvps that I use to show as possible and working, and then that gets sent to engineering to convert to something that follows better best practices.

This is what I use and like:

https://yourls.org

Same, Gimp and Inkscape have been more than adequate for everything I need.

Same, they had some kind of truffle pizza that was delicious

I assumed this was because it’s internal devs that are actively looking for platforms to build on, while external devs might already have stacks they prefer to work on due to how it interacts with their security postures. Pure speculation, but it’s been my anecdotal experience.

I recently upgraded my USG for a dream machine, glad it seems the upgrade was worth it.

I use n8n for managing ChatGPT's API and to connect it to other APIs like gmail. It helps abstract away the authentication side of APIs which can be very cumbersome to manage.

It also helped me better understand the chatgpt API. N8n allows you to code in raw javascript or python which has allowed me to branch out to pure python.

I would say that no-code low code has a place as a stepping stone, that can allow other invested parties besides developers to manage inter departmental flow and build a blueprint that can be redone with actual engineers in the preferred manner.

Isn’t that where social media is consumed? It makes sense that this isolation drives us to connect, and in ways that are not as helpful as real meaningful connections.

That’s a great analogy thank you!

I agree, this is exactly my experience as well. My earlier requests for code were more thorough then they are now. I can still use it if I give it the documentation in the prompt, but it has trouble remembering context across prompts.

I see it differently. They built the best they could, but it was expensive. Now it feels to me that they are "optimizing" it for performance by seeing if they can lower the quality but still give people just enough value to pay for it. Makes me think they are trying to find a sweet spot between performance and cost. I am seeing this drop in quality, but mostly when it comes to the model remembering things you've told it in the past, and a much bigger dependence on custom instructions for prompts.

I still get value, but I have to give it the documentation I want it to focus on during the prompt, instead of referencing it earlier in the conversation.

This makes me think they are lowering it's ability to store context so that the calls don't cost as much to make.

Forgive my conjecture, as I'm no expert in this field, but have enjoyed playing with the different models.

I see it as a matter of attention. GPT-4 is limited in the number of tokens you can feed it and receive back from it. I see this as how much attention the LLM can give you. I see meta-gpt and other models like this as increasing the attention of the LLM by allowing it combine it's short attention span with multiple assessments of the request, and give you a more complete picture of what you want because it can keep giving it consistent attention to the problem at hand, instead of simply trying to solve it on the first try.

Someone please correct me if I'm way off, but this simple mental model helps me abstract this.

This is how I feel. I chose to have children with the knowledge that there is a very real possibility that they may inherit a hell scape when I’m dead. The thought of them having the possibility of a better future combined with finally having something positive to hope for fills a void that I believe many people share.

Agreed, I just attempted it via the api and Ui and it gave a mishmash response that didn’t seem to be coherent, mostly different subjects all smashed together at the max length output.

I tried it via api two ways, both via system message and no user message and one with the word rewrite at the beginning and one without. Both gave similar long mishmashed outputs that seemed like a hallucination.

I believe this blog post has a breakdown of all the updates and the twitter posts with the videos: https://eirifu.wordpress.com/2023/07/30/lk-99-superconductor...

I wonder if this is a better use case for “smart” eyeglasses? Audio as the input at first, have the audio files sync wirelessly to your phone, and apply the ML transcription and prompt keys locally.

I believe management has in many cases lost focus on strategy because they are either not driven as well since they also are not fully invested in the company’s overall vision or are so overwhelmed with meetings that they simply don’t have time to even understand what their subordinates actually do in technical terms, much less strategize.

I think the distinction being drawn here is that there is a disconnect between hiring managers and workers.

A hiring manager has an expectation that 200k will incentivize someone to work for 8 hours a day consistently and constantly where work is the point, not the tasks themselves. They expect the worker to figure out what additional tasks are to be done for that price and work at the upper limit of the time expectations (8 hour a day).

Whereas the average worker is coming to the workplace to complete a set of tasks defined by the manager, not to figure out more tasks to do in their free time. They expect the manager to know what tasks needs to be done and understand that their results need to be tracked.

As jobs become more and more compartmentalized it becomes less of an advantage for an employee to be hungry for more work, when their chance for promotion diminishes the better they are at their job, as in my anecdotal experience, management is loathe to lose that efficient cog in the system in that specific point. Which then reinforces the employees becoming jaded and resorting to the least amount of work possible.

I believe it’s because in todays corporate world more often than not, most employees have no skin in the game. That is, they are not privy to the more long term planning for their companies and are not usually compensated based on performance (outside of sales anyway).

I agree with what you are saying. Much like a muscle, when you work out, you're not lifting weights so you can lift more weights later.

Your general strength is higher, so most physical tasks are easier. Same for reading, if you practice reading books and enjoy looking at the written word, then reading developer documentation doesn't seem as intimidating which opens more doors for you.

Depends on what I’m writing. For anything vanilla I use the chatgpt site Ui. For anything that maybe censored like smut I use prompt engineering and gpt 3.5 turbo via api to get past the censors, although they just announced this model is to be deprecated in January so I’m playing with local language models to some success. For now been using gpt4all and it’s surprisingly good, although slow on my hardware.

I'm similar, and I've found AI to help with this. I'm able to guide the AI to write the first draft, then I go in and rewrite it. It's helped remove that frustration at the first draft.

I use the iOS shortcut for pasting into Joplin and it works great other than not brining in the title.

Wanted to thank you for the great app! I left Evernote a year ago and it’s helped me get my journaling back on track. I use Dropbox sync and it wonderfully. I don’t remember the last time I had issues with it.

Only thing I hope for one day is the ability to see the article tags on each note in the list view so I can see which notes have which tags in common without having to click on the tag.

Keep up the great work!

I would love to know this as well. Can the current automation and robotics driven systems make this a more efficient and viable option for manufacturing?

That is why control is so important. Just like Wordpress took over the web design landscape, an open source model you control at home is important in countering the specter of the big bro ai.