Now working on reading news feeds from websites #buildinpublic #software-dev
One Two Bytes
@onetwobytes.bsky.social
Roaming the software world. A smorgasbord of topics I come across while writing software. https://www.onetwobytes.com
Added a reader mode to the agent webloader, showing good results in extracting meaningful content from cluttered websites #buildinpublic #software-dev
Problem 2: Big context windows ≠ smarter LLMs often drew incorrect conclusions or overlooked details when parsing the entire HTML for the page. We had to transform everything into simplified, structured formats like CSV with filtered columns.
Build a local agent for privacy and control. No cloud. No API Keys. #SoftwareDev #LocalAI #OpenSource #Ollama #Agents
Build Your Own Local AI Agent: A Step By Step Guide
Build fast, private AI agents that run fully on your machine. No cloud, no API keys. In this post, I'll show how to set up a local agent that scrapes news headlines and summarizes them using open-source tools like LangChain and Ollama.
onetwobytes.com
Wrote a OneTenAgent. Example: one_ten_agent.run("On a scale from 1 to 10, how interesting would this post be for Swimmers?") and it returns the number. Works surprisingly well github.com/tm243/agent-... #PythonDev #SoftwareDev
Chain-of-Thought Prompting guide for complex reasoning: www.promptingguide.ai/techniques/cot #Artificiallntelligence #AI #RAG #AgenticAl #AlagentInnovation #AlAgents #LLMs #LLM #GenerativeAl #GenAl #technology #TechRevolution #tech #Engineering
Chain-of-Thought Prompting – Nextra
A Comprehensive Overview of Prompt Engineering
promptingguide.ai
it's possible to run a small LLM like Gemma3 4b using Ollama in a basic CircleCI pipeline #Dev #PythonDev #LLM
How to run Ollama in CircleCI
It's possible to run an LLM like Gemma3 4b using Ollama in a basic CircleCI pipeline, which allows you to integrate lightweight AI capabilities, such as semantic unittests or simple agents, into your CI/CD workflows.
onetwobytes.com
I also think one of the most powerful forms of radicalization is educating as many people as possible about mechanics - financial, systemic and technological - behind the tech industry. Their entire success has been based on obfuscating their work as too big or complex for "normal" people to "get."
Pluses of using LLMs in coding: - Quickly showcase prototypes and get instant client feedback. Minuses of using LLMs in coding: - Explaining to clients that prototypes still require proper development before production.
Ollama Docker image with Gemma3:4b works well on CircleCI’s free plan. #dev #PythonDev #LLM
Typescript 7 compiler will be written in Go devblogs.microsoft.com/typescript/t...
Just wrote about how traditional unit tests struggle with slightly changing text and using semantic assertions to test meaning instead - practical for LLM output. #unittest #LLM #PythonDev #Development #GameDev
Semantic Unittests
Traditional unit tests struggle with AI-generated text that varies slightly each time. Semantic unit testing solves this by comparing meaning rather than exact wording, making it ideal for validating AI outputs like classification, summarization, and text moderation.
onetwobytes.com
In 2024, Python surpassed JavaScript as the most popular programming language on GitHub for the first time. #programming #coding #python #java #javascript #typescript #csharp #c++ #php #c #go #golang #shell #ruby #obj-c
VS Code 1.98 Copilot Vision supposedly writes HTML/CSS based on a screenshot, has anyone tried that? #dev
Gotta check how a TrapC program integrates with existing C and C++ codebases. #C #C++ #Programming #MemorySafety #Tech
reading the source code of Command & Conquer Tiberian Dawn today ☺️ #OpenSource #GameDev #RetroGaming #CodeAsLiterature #DigitalArchaeology
Working with Copilot Just something I wanted to share: Me: when working with Copilot, many things get easier, but the core problems remain the same and fixing hairy problems such as concurrency issues and the like, or test edge cases etc are still manual work. It helps mainly with boilerplate…
Working with Copilot
Just something I wanted to share: Me: when working with Copilot, many things get easier, but the core problems remain the same and fixing hairy problems such as concurrency issues and the like, or test edge cases etc are still manual work. It helps mainly with boilerplate code, which aren't difficult to write anyway, but it saves time ChatGPT said:
onetwobytes.com
Vertical AI agents are transforming the AI industry Here’s what you need to learn to put yourself out there. • Understand Autonomous Agents • Master Goal-Driven Behavior • Explore Multi-Agent Systems • Learn Reinforcement Learning Basics • Hands-On Practice
NotOpenSSLWarning: urllib3 v2 only supports OpenSSL 1.1.1+, currently the ‘ssl’ module is compiled with ‘LibreSSL 2.8.3’. site-packages/urllib3/__init__.py:34: NotOpenSSLWarning: urllib3 v2 only supports OpenSSL 1.1.1+, currently the 'ssl' module is compiled with 'LibreSSL 2.8.3'. See: Turns out…
NotOpenSSLWarning: urllib3 v2 only supports OpenSSL 1.1.1+, currently the ‘ssl’ module is compiled with ‘LibreSSL 2.8.3’.
site-packages/urllib3/__init__.py:34: NotOpenSSLWarning: urllib3 v2 only supports OpenSSL 1.1.1+, currently the 'ssl' module is compiled with 'LibreSSL 2.8.3'. See: Turns out that urllib3 version 2 wants OpenSSL to work properly. The earlier versions of urllib3 version 2 even stopped entirely, the later ones throw this warning. But my current ssl module appears to be LibreSSL. The idea is to install an…
onetwobytes.com
How to install the yaml package for Python? I want to read the config for my backend from a yaml file and when installing the yaml package I am getting the following error: pip install yaml Looking in indexes: ERROR: Could not find a version that satisfies the requirement yaml (from versions:…
How to install the yaml package for Python?
I want to read the config for my backend from a yaml file and when installing the yaml package I am getting the following error: pip install yaml Looking in indexes: ERROR: Could not find a version that satisfies the requirement yaml (from versions: none) ERROR: No matching distribution found for yaml Quick Answer pip install pyyaml More Details…
onetwobytes.com
How to Install and Use Salesforce’s CodeGen LLM CodeGen is an AI (LLM) from Salesforce that can generate source code, as well as describe what a piece of code does. It comes under the Apache license and has a good performance while being lightweight enough to run on a laptop for both inference and…
How to Install and Use Salesforce’s CodeGen LLM
CodeGen is an AI (LLM) from Salesforce that can generate source code, as well as describe what a piece of code does. It comes under the Apache license and has a good performance while being lightweight enough to run on a laptop for both inference and fine tuning. Here is how to set it up and how to use it. …
onetwobytes.com
Python-Alpaca Dataset A dataset of 22,600 verified Python code examples. Tested for functionality and sourced from top projects, it’s perfect for training AI models on reliable, clean code.
Python-Alpaca Dataset
A dataset of 22,600 verified Python code examples. Tested for functionality and sourced from top projects, it’s perfect for training AI models on reliable, clean code.
onetwobytes.wordpress.com
Hey Bluesky! I’m looking for a #software engineer! Being able to build a platform (cloud services, frameworks, APIs, etc) from scratch is a must. 🌟🤸🏼♀️ #hiring #softwareengineer #tech
At Netlify we've been building specific flows and features for the new AI agents building on and deploying to our product. Today I'm introducing the term AX, agent experience, that gives name to a concept of great importance to software development going forward. biilmann.blog/articles/int...
Introducing AX: Why Agent Experience Matters
As builders, we need to start focusing on AX or “agent experience” — the holistic experience AI Agents will have as the user of a product or platform.
biilmann.blog
#Localfirst is about giving data ownership and agency to end-users while making the #Cloud optional. @adamwiggins.bsky.social @expede.wtf @biilmann.blog @yonz.bsky.social explain why this emerging model for #software #development could change everything. www.heavybit.com/library/arti...
AFAIK the easiest way to install arbitrary “self-hosted” AI models on a Mac is with an app/command-line-tool called ollama: ollama.com … and there are apps like Chital (https://github.com/sheshbabu/Chital) that improve the UX of interacting with the ollama “server” on your Mac.