Amazon's new Ocelot chip brings us closer to building a practical quantum computer
New 'Ocelot' chip uses scalable architecture for reducing error correction by up to 90% and accelerating the development of real-world quantum computing applications.
Wednesday, March 12, 2025
Tuesday, March 11, 2025
Elon Musk's Algorithm: 5-Step Process
Elon Musk's 'Algorithm,' a 5-Step Process to Dramatically Improve Nearly Anything, Is Both Simple and Brilliant
- Question every requirement. ...
- Delete any part of the process you can. ...
- Simplify and optimize. ...
- Accelerate cycle time. ...
- Automate.
from book:
Monday, March 10, 2025
Agentic AI: Microsoft AutoGen vs LangGraph vs CrewAI
Which Agentic AI Framework to Pick? LangGraph vs. CrewAI vs. AutoGen - YouTube
whitew1994WW (Will White) · GitHub
AI Trends 2025: AI Agents and Multi-Agent Systems with Victor Dibia | The TWIML AI Podcast
- AI Agents 2024 Rewind - A Year of Building and Learning
- AutoGen @ GitHub
- AutoGen v0.4: Reimagining the foundation of agentic AI for scale, extensibility, and robustness
- Multi-Agent Systems with Autogen Book
- Magentic-One: A Generalist Multi-Agent System for Solving Complex Tasks
- GAIA: a benchmark for General AI Assistants
- OmniParser: Screen Parsing tool for Pure Vision Based GUI Agent
- LangGraph
- LlamaIndex
- ByteDance UI-TARS
- The Human-AI eXperiences (HAX) Team
- AI Engineering Pitfalls with Chip Huyen - #715
- Why Agents Are Stupid & What We Can Do About It with Dan Jeffries - #713
- Why Your RAG System Is Broken, and How to Fix It with Jason Liu - #709
Sunday, March 09, 2025
programming puzzles
for training humans and AI to code...
Project Euler is a series of challenging mathematical/computer programming problems that will require more than just mathematical insights to solve. Although mathematics will help you arrive at elegant and efficient methods, the use of a computer and programming skills will be required to solve most problems.The motivation for starting Project Euler, and its continuation, is to provide a platform for the inquiring mind to delve into unfamiliar areas and learn new concepts in a fun and recreational context.
Nifty Assignments @stanford.edu
The Nifty Assignments session at the annual SIGCSE meeting is all about gathering and distributing great assignment ideas and their materials. For each assignment, the web pages linked below describe the assignment and provides materials — handouts, starter code, and so on.For instance, humuhumunukunukuapua’a phonetic guide would be Hoo-moo-hoo-moo-noo-koonoo-koo-ah-poo-ah'ah
The 12 valid characters in the Hawaiian language are a, e, i, o, u, p, k, h, l, m, n, and w.
Knuth: Selected Papers on Fun and Games @ Stanford
Saturday, March 08, 2025
AI: LangChain, LangGraph and Agentic AI
podcast
LangChain and Agentic AI Engineering with Erick Friis - Software Engineering Daily
LangChain.com
Tutorials | 🦜️🔗 LangChain (Python)
GitHub - langchain-ai/langchain: 🦜🔗 Build context-aware reasoning applications
Tutorials | 🦜️🔗 Langchain (JavaScript)
GitHub - langchain-ai/langchainjs: 🦜🔗 Build context-aware reasoning applications 🦜🔗
LangGraph Gain control with LangGraph to design agents that reliably handle complex tasks. Build and scale agentic applications with LangGraph Platform.
Friday, March 07, 2025
EV mini-max truck: Telo
ONE TRUCK TO RULE THEM ALL... - YouTube
The Telo mini Electric truck solves a lot of problems for truck owners, especially if you live in a city. With 500 horse power, 10 inches of ground clearance, and 6600 pounds of towing capacity the Telo truck can also handle life off of the paved trails.full 6-8' truck bed, $41K
Thursday, March 06, 2025
Data formats: Apache Parquet vs ORC vs Avro
Why Parquet vs. ORC: An In-depth Comparison of File Formats | by Ankush Singh | Medium
Apache Parquet is a columnar storage file format available to any project in the Hadoop ecosystem. It’s designed for efficiency and performance, and it’s particularly well-suited for running complex queries on large datasets.
Parquet is an excellent choice when dealing with large, complex, and nested data structures, especially for read-heavy workloads or when you want to perform analytics using tools like Apache Spark or Apache Arrow. Its columnar storage approach makes it an excellent choice for data warehousing solutions where aggregation queries are common.
Parquet
while Parquet is a columnar data format, for "row" level storage is is often used "Avro", that besides efficiently storing data in binary format also includes a data schema.
Apache Parquet is a columnar storage file format available to any project in the Hadoop ecosystem. It’s designed for efficiency and performance, and it’s particularly well-suited for running complex queries on large datasets.
Parquet is an excellent choice when dealing with large, complex, and nested data structures, especially for read-heavy workloads or when you want to perform analytics using tools like Apache Spark or Apache Arrow. Its columnar storage approach makes it an excellent choice for data warehousing solutions where aggregation queries are common.
Parquet
Apache Parquet - Wikipedia
GitHub - apache/parquet-format: Apache Parquet Format
ORC is another popular file format in the Hadoop ecosystem. It’s a self-describing, type-aware columnar file format designed for Hadoop workloads.
ORC is commonly used in cases where high-speed writing is necessary, particularly with Hive-based frameworks. It also suits well when data modifications (updates and deletes) are needed in your use case because it supports ACID properties. Lastly, ORC is a good choice when using complex and nested data types.
GitHub - apache/parquet-format: Apache Parquet Format
ORC is another popular file format in the Hadoop ecosystem. It’s a self-describing, type-aware columnar file format designed for Hadoop workloads.
ORC is commonly used in cases where high-speed writing is necessary, particularly with Hive-based frameworks. It also suits well when data modifications (updates and deletes) are needed in your use case because it supports ACID properties. Lastly, ORC is a good choice when using complex and nested data types.
while Parquet is a columnar data format, for "row" level storage is is often used "Avro", that besides efficiently storing data in binary format also includes a data schema.
Apache Avro - Wikipedia
Avro is a row-oriented remote procedure call and data serialization framework developed within Apache's Hadoop project. It uses JSON for defining data types and protocols, and serializes data in a compact binary format. Its primary use is in Apache Hadoop, where it can provide both a serialization format for persistent data, and a wire format for communication between Hadoop nodes, and from client programs to the Hadoop services. Avro uses a schema to structure the data that is being encoded.
Avro relies on schemas. When Avro data is read, the schema used when writing it is always present. This permits each datum to be written with no per-value overheads, making serialization both fast and small. This also facilitates use with dynamic, scripting languages, since data, together with its schema, is fully self-describing.
Wednesday, March 05, 2025
OpenAI: $20,000/month for AI 'agents' ?
OpenAI reportedly plans to charge up to $20,000 a month for specialized AI 'agents' | TechCrunch
OpenAI may be planning to charge up to $20,000 per month for specialized AI “agents,” according to The Information.
The publication reports that OpenAI intends to launch several “agent” products tailored for different applications, including sorting and ranking sales leads and software engineering. One, a “high-income knowledge worker” agent, will reportedly be priced at $2,000 a month.
Another, a software developer agent, is said to cost $10,000 a month.
...$20,000-per-month ...“PhD-level research,”
...$20,000-per-month ...“PhD-level research,”
Tuesday, March 04, 2025
book: Generative AI on AWS
- Apply generative AI to your business use cases
- Determine which generative AI models are best suited to your task
- Perform prompt engineering and in-context learning
- Fine-tune generative AI models on your datasets with low-rank adaptation (LoRA)
- Align generative AI models to human values with reinforcement learning from human feedback (RLHF)
- Augment your model with retrieval-augmented generation (RAG)
- Explore libraries such as LangChain and ReAct to develop agents and actions
- Build generative AI applications with Amazon Bedrock
Monday, March 03, 2025
Visual AI (ANN)
𝗼𝗻𝗲 𝗼𝗳 𝘁𝗵𝗲 𝗕𝗘𝗦𝗧 𝘃𝗶𝘀𝘂𝗮𝗹𝗶𝘇𝗮𝘁𝗶𝗼𝗻𝘀 𝗼𝗳 𝗵𝗼𝘄 𝗮 𝗡𝗲𝘂𝗿𝗮𝗹 𝗡𝗲𝘁 𝘄𝗼𝗿𝗸𝘀 @LinkedIn
A neural net is made up of three key components:
🔹 Artificial Neurons (Nodes) – Basic computing units that process and transmit information.
🔹 Connections (Synapses) – Links between neurons, each with an associated weight that determines signal strength.
🔹 Activation Functions – Mathematical functions that decide which neurons get activated and contribute to the final prediction.
The more neurons and connections a neural net has, the more complex patterns it can learn:
- Higher accuracy – More neurons = better pattern recognition.
- Increased capacity – The network can store and recall more intricate relationships.
🔹 Artificial Neurons (Nodes) – Basic computing units that process and transmit information.
🔹 Connections (Synapses) – Links between neurons, each with an associated weight that determines signal strength.
🔹 Activation Functions – Mathematical functions that decide which neurons get activated and contribute to the final prediction.
The more neurons and connections a neural net has, the more complex patterns it can learn:
- Higher accuracy – More neurons = better pattern recognition.
- Increased capacity – The network can store and recall more intricate relationships.
Sunday, March 02, 2025
Saturday, March 01, 2025
AI agents "language": GibberLink
TypeScript, React, OpenAI, 11labs
gbrl.ai — Agent2Agent conversation in your browser (use two devices)Two independent conversational ElevenLabs AI agents are prompted to chat about booking a hotel (one as a caller, one as a receptionist)Both agents are prompted to switch to ggwave data-over-sound protocol when they identify other side as AI, and keep speaking in english otherwise
This repository provides API that allows agents to use the protocol
Bonus: you can open the ggwave web demo, play the video above and see all the messages decoded!
This mode involves a series of sounds that combine the reminiscent tones of dial-up modems and the beeps of R2-D2 from Star Wars, creating a communication method incomprehensible to human listeners.
Bonus: you can open the ggwave web demo, play the video above and see all the messages decoded!
This mode involves a series of sounds that combine the reminiscent tones of dial-up modems and the beeps of R2-D2 from Star Wars, creating a communication method incomprehensible to human listeners.
Open-Source AI for Biology at MIT
Developed by a team of researchers in the MIT Jameel Clinic for Machine Learning in Health, Boltz-1 is the first fully open-source model that achieves state-of-the-art performance at the level of AlphaFold3, the model from Google DeepMind that predicts the 3D structures of proteins and other biological molecules.
Generative AI imagines new protein structures | MIT News | Massachusetts Institute of Technology
AI used to design a multi-step enzyme that can digest some plastics - Ars Technica