_pdp_

25 ms·

451.
by _pdp_ 10mo ago
AI agent reference architecture that utilizes a collection of playbooks to solve specific problems an a structured and predictable way. The architecture mimics real human systems based on checklists and playbooks.
452.
by _pdp_ 10mo ago
There is no yes and no. This is a fact. Even a small startup of 3-5 people will pay more in terms of salaries than the total license costs they consume. A larger enterprise will will spend 50 to 100 times more on salaries then software lice
453.
by _pdp_ 10mo ago
This is only true if you assume that you are producing the same amount of code as today. Though, AI ultimately will produce more code which will require higher maintenance. Your internal team will need to scale up due to the the amount of c
454.
by _pdp_ 10mo ago
I often give the follow analogy which I think is a good proxy to what is going on. Spreadsheets! They are everywhere. In fact, they are so abundant these days that that many are spawned for a quick job and immediately discarded. In fact, th
455.
by _pdp_ 10mo ago
https://chatbotkit.com/examples/playbook-based-agent-referen...
456.
by _pdp_ 10mo ago
In my previous company we did everything as a micro service. In the company before that it was serverless on AWS! In both cases we had to come up with clever solutions to simply get by because communication between services is of a problem.
457.
by _pdp_ 10mo ago
We ended up using JSONL for our implementation but it is also possible to switch using other "streamable" variants like SSE and even CSV if that is what you are into. But, as the OP suggested, none of these are resumable. So we ha
458.
by _pdp_ 10mo ago
No. You basically call a function to temporarily or permanently extend the base prompt. But of course you can think of other patterns to do more interesting things depending on your use-case. The prompt selection is a RAG.
459.
by _pdp_ 10mo ago
This is exactly what I do. It works super well. Who would have thought that documenting your code helps both other developers and AI agent? I've been sarcastic.
460.
by _pdp_ 10mo ago
Oh nice. I did not know. Thanks for the link.
461.
by _pdp_ 10mo ago
???
462.
by _pdp_ 10mo ago
It is not an invention if it is common sense and there is plenty of previous art. How would you otherwise dynamically extend the prompt? You will have some kind of function that based on the selected preferences add more prompt to the base
463.
by _pdp_ 10mo ago
They exist if you are worth 70 billion.
464.
by _pdp_ 10mo ago
It is the same. Context7 is using LLMs.txt to create a searchable index that can be used for coding and Q&A. It serves the same purpose as this tool except I guess it is more standard if you even call it that at this stage.
465.
by _pdp_ 10mo ago
I was thinking the same. While it is not for everyone, hacking the TV to make the dumb is possible.
466.
by _pdp_ 10mo ago
I hate to be that guy but skills are not an invention of sorts. It a simple mechanism that exists already in many places. The biggest unlock was tool calling that was in invented at OpenAI.
467.
by _pdp_ 10mo ago
LLMs need prompts. Prompts can get very big very quickly. The so called "skills", which exist in other forms in other platforms outside of Anthropic and OpenAI, are simply a mechanism to extend the prompt dynamically. The tool (sc
468.
by _pdp_ 10mo ago
Fun idea although I thought the industry is leaning towards using llms.txt.
469.
by _pdp_ 10mo ago
What is the difference between this and let's say Claude Code using something like semgrep as a tool? Also I don't think this tool should be in the developer flow as in my experience it is unlikely to run it on the regular. It sho
470.
by _pdp_ 11mo ago
Many of us grew up without social media and turned out to be fine! Anyone arguing against this ban is delusional what social media does to undeveloped brains. There are plenty of studies to support this as well. Social media is harmful to c
471.
by _pdp_ 11mo ago
The majority of vibe coded software I have seen (and produced) are similar to spreadsheets - it does the job for 5 minutes - easily forgettable. It is throwaway software at best. What AI did was to move the goalpost even further. Now you ne
472.
by _pdp_ 11mo ago
So no static compiler checks and apparently no fuzzers used to ensure these rules work as intended?
473.
by _pdp_ 11mo ago
I don't know what the LLM said is true for sure but based on my experience in the field sounds plausible. The only way to know is to verify it. Btw, LLMs are already used in vulnerability discovery and exploit development.
474.
by _pdp_ 11mo ago
An operator might not be able to upgrade at all! Along the fixes, the advisories now need to contain detailed workarouds, firewall rules and other adhoc solutions to ensure they get quickly deployed.
475.
by _pdp_ 11mo ago
I don't have time to look into it right now (def later)! However, I was curious to see if github copilot can reverse engineer it based on the latest commits and seems that what it is saying aligns with both advisories. It pointed out t
476.
by _pdp_ 11mo ago
It makes total sense to me.
477.
by _pdp_ 11mo ago
I am not surprised at all. I can already see self improving behaviour in our own work which means that the next logic step is self improving! I know how this sounds but it seems to me, at least from my own vantage point, that things are mov
478.
by _pdp_ 11mo ago
You put a warning where it is most likely to be seen by a human coder. Besides, no amount of prompting will prevent this situation. If it is a concern then you put a linter or unit tests to prevent it altogether, or make a wrapper around th
479.
by _pdp_ 11mo ago
There is far much easier way to do this and one that is perfectly aligned with how these tools work. It is called documenting your code! Just write what this file is supposed to do in a clear concise way. It acts as a prompt, it provides mu
480.
by _pdp_ 11mo ago
Skills are fundamentally different. The only similarity I can see is perhaps the fact that in both cases an LLM is used to find the right skillset.

More ›