Tuesday, December 31, 2024

A2I: Amazon Augmented AI

Machine Learning Workflow - Amazon Augmented AI - AWS

Amazon Augmented AI (Amazon A2I) allows you to conduct a human review of machine learning (ML) systems to guarantee precision.


Amazon A2I features – Amazon Web Services

Course: [NEW] Ultimate AWS Certified AI Practitioner AIF-C01 | Udemy Business


Posted by at No comments:

Monday, December 30, 2024

AI: LWMs: World Labs, 3D virtual worlds

Generating Worlds

World Labs: a spatial intelligence AI company building Large World Models (LWMs) to perceive, generate, and interact with the 3D world. We aim to lift AI models from the 2D plane of pixels to full 3D worlds - both virtual and real - endowing them with spatial intelligence as rich as our own.

World Labs was founded by visionary AI pioneer Fei-Fei Li along with Justin Johnson, Christoph Lassner, and Ben Mildenhall; each a world-renowned technologist in computer vision and graphics.

Hello, World Labs




similar:


Genie 2, a foundation world model capable of generating an endless variety of action-controllable, playable 3D environments for training and evaluating embodied agents. Based on a single prompt image, it can be played by a human or AI agent using keyboard and mouse inputs.
Posted by at No comments:

Sunday, December 29, 2024

AGI defined: $100B profit (OpenAI, Microsoft)

Microsoft and OpenAI have a financial definition of AGI: Report | TechCrunch

"The two companies reportedly signed an agreement last year stating OpenAI has only achieved AGI when it develops AI systems that can generate at least $100 billion in profits. That’s far from the rigorous technical and philosophical definition of AGI many expect."



Artificial general intelligence (AGI) is a type of artificial intelligence (AI) that matches or surpasses human cognitive capabilities across a wide range of cognitive tasks. This contrasts with narrow AI, which is limited to specific tasks.[1] Artificial superintelligence (ASI), on the other hand, refers to AGI that greatly exceeds human cognitive capabilities. AGI is considered one of the definitions of strong AI.



Posted by at No comments:

tool: PGlite, Posgres DB in WASM

PGlite.dev

Embeddable Postgres

Run a full Postgres database locally in WASM with reactivity and live sync.


Get Started

GitHub

What is PGlite | PGlite

PGlite is a WASM Postgres build packaged into a TypeScript/JavaScript client library, that enables you to run Postgres in the browser, Node.js and Bun, with no need to install any other dependencies. It's under 3mb Gzipped, and has support for many Postgres extensions, including pgvector.





Posted by at No comments:

Saturday, December 28, 2024

Calendar 2025: AI Serbia v3

continuous calendar 2025 "months" (new)
with AI illustrations in impressionist style (new)
Serbia (new)
with national holidays,
and moon phases (new)

printable PDFs


Serbian, ћирилица, holidays, moon phases, A4 size

English, US, Letter size

English, US official holidays, moon phases (EST), Letter size



"seasons" (winter, spring, summer, autumn/fall)





DraganSr: Calendar 2025: AI National Parks, v3,
months & seasons
Posted by at No comments:

tool: electric-sql for Postgres DB

 electric-sql/electric: Sync little subsets of your Postgres data into local apps and services. @GitHub

Electric provides an HTTP API for syncing Shapes of data from Postgres. This can be used directly or via client libraries and integrations.

Quickstart | ElectricSQL



Shapes - Guide | ElectricSQL

Electric syncs little subsets of your Postgres data into local apps and services. Those subsets are defined using Shapes. A shape is a way of defining a subset of that data that you'd like to sync into a local app. Defining shapes allows you to sync just the data you want and just the data that's practical to sync onto the local device.




Posted by at No comments:

Friday, December 27, 2024

AI: Digital Biology

Google Deepmind CEO's STUNNING Prediction - Digital Biology - YouTube

Demis Hassabis - Wikipedia

Sir Demis Hassabis (born 27 July 1976) is a British artificial intelligence (AI) researcher, and entrepreneur. He is the chief executive officer and co-founder of Google DeepMind,[8] and Isomorphic Labs,[9][10][11] and a UK Government AI Adviser.[12] In 2024, Hassabis and John M. Jumper were jointly awarded the Nobel Prize in Chemistry for their AI research contributions for protein structure prediction.

Demis Hassabis – Nobel Prize lecture - NobelPrize.org

Nobel Prize lecture: Demis Hassabis, Nobel Prize in Chemistry 2024 - YouTube

Accelerating scientific discovery with AI (PDF)

AlphaFold: a solution to a 50-year-old grand challenge in biology - Google DeepMind


Digital Biology | Andreessen Horowitz


Daphne Koller is the founder and CEO of insitro, a company using AI and machine learning to engineer drug discovery. In this conversation with a16z's Vijay Pande, she discusses how an LLM for cells could revolutionize drug discovery and how to bridge the technological and cultural divide between building with atoms and building with bits.


Daphne Koller (born August 27, 1968) is an Israeli-American computer scientist. She was a professor in the department of computer science at Stanford University[4] and a MacArthur Foundation fellowship recipient.[1] She is one of the founders of Coursera, an online education platform. Her general research area is artificial intelligence[5][6] and its applications in the biomedical sciences.


Posted by at No comments:

tool: ArchiveBox

 Let's archive the web with Nick Sweeting from ArchiveBox (Changelog Interviews #619)



Posted by at No comments:

Thursday, December 26, 2024

EV: Honda += Nissan

Honda, Nissan Confirm Plans to Merge into Third-Largest Automaker @CarAndDriver

  • Honda and Nissan confirmed discussions between the two companies that would form the third-largest automaker in the world based on sales.

  • the goal of completing the merger by August 2026

  • jointly developing electric vehicles and automotive software

  • Mitsubishi Motors' participation







Posted by at No comments:

AI: multi-modal embeddings model @ Amazon Bedrock

Amazon Titan is "native" multi-modal (text, images) AI model trained by and available at AWS


Using Titan Multimodal Embeddings, you can generate embeddings for your content and store them in a vector database. When an end user submits any combination of text and image as a search query, the model generates embeddings for the search query and matches them to the stored embeddings to provide relevant search and recommendations results to end users. For example, a stock photography company with hundreds of millions of images can use the model to power its search functionality, so users can search for images using a phrase, image, or a combination of image and text.




Posted by at No comments:

Local-first web apps

 Local-first, y/n? with Johannes Schickling & James Long (Changelog & Friends #71)


Posted by at No comments:

Wednesday, December 25, 2024

AWS AI Certification

AWS Certification - Validate AWS Cloud Skills - Get AWS Certified

AWS-Certified-AI-Practitioner_Exam-Guide.pdf

Posted by at No comments:

Quantum Internet

 What is the quantum internet? | University of Chicago News

The quantum internet is a network of quantum computers that will someday send, compute, and receive information encoded in quantum states.

The quantum internet will not replace the modern or “classical” internet; instead,
it will provide new functionalities such as quantum cryptography and quantum cloud computing.



Quantum internet utilizes quantum bits (qubits) instead of classical binary bits. The fascinating property of qubits is that the mere act of observing them changes their state – making it impossible for hackers to intercept messages without detection. This isn’t theoretical physics; it’s happening now. Countries like China already have satellites in orbit sending quantum-encrypted messages.



Bernardo Huberman is a fellow and vice president of the Next-Gen Systems Team at CableLabs.[1] He is also a consulting professor in the Department of Applied Physics and the Symbolic System Program at Stanford University.



It involves the sharing of information at atomic and subatomic levels through quantum channels. Compared to classical computing, it can share information at infinitely higher rates and with vastly fewer limitations. It is also a great deal more secure than classical computing.






Posted by at No comments:

Tuesday, December 24, 2024

tk9.0: GoLang cross-platform GUI

 tk9_0 package - modernc.org/tk9.0 - Go Packages

The CGo-free, cross platform GUI toolkit for Go



Posted by at No comments:

AI HW: NPU vs GPU; TOPS


Efficiency: While GPUs are capable of handling many demanding tasks associated with AI, NPUs are purpose-built for these requests and can meet similar or even better performance benchmarks while requiring exponentially less power, a particularly valuable feature for battery-powered devices with finite capacity.

CPU vs GPU vs NPU: What's the difference? | CORSAIR

Neural Processing Units (NPUs)...are designed to efficiently handle matrix multiplication and addition, which is essential for artificial intelligence (AI) and machine learning (ML) workloads
such as image recognition, natural language processing, and machine learning."


TOPS quantifies an NPU's processing capabilities
by measuring the number of operations (additions, multiplies, etc.)
in trillions executed within a second.


An AI accelerator, deep learning processor or neural processing unit (NPU) is a class of specialized hardware accelerator[1] or computer system[2][3] designed to accelerate artificial intelligence and machine learning applications

NPU uses less power and is far more efficient at AI tasks than a CPU or GPU.



micro "neural network" example, image detection with matrix multiplication


A specialized NPU does not take up much space on processor chips.




$499, 38 TOPS; a great "deal" for this price

$249, 67 TOPS, "on backorder"





Posted by at No comments:

Monday, December 23, 2024

AI Agents vs Software Apps

Microsoft CEO’s Shocking Prediction: “Agents Will Replace ALL Software" - YouTube

"business application (CRUD) will collapse in AI agents era"

"AI (agents/model) will be the place where all logic is"

Copilot is UI for AI; instead of SW apps; just need a database


Satya Nadella | BG2 w/ Bill Gurley & Brad Gerstner - YouTube (BG2 podcast)


Microsoft CEO Satya Nadella Highlights AI Agent Capabilities Across Microsoft Products - YouTube





Posted by at No comments:

BOXABL: Casita, $60K "Lego" House

multiple "boxes" can be assembled in a larger house: x2, x3, x4
one "box" was $50K, now $60K



  • Affordable: High-quality living without the high price tag.
  • Compact & Stylish: 361 sq. ft. of modern design, including a full-sized kitchen, bathroom, and living space.
  • Quick & Easy Setup:
  • Unfolds in 1 hour and sets up in a day.
  • Sustainable: Built with energy efficiency in mind to lower your environmental footprint.
  • 19ft x 19ft
  • 9’6” ceilings
  • 361 sqft
  • 6ft Windows
  • 8ft Doors
  • Vinyl Flooring





Posted by at No comments:

Sunday, December 22, 2024

Calendar 2025: AI National Parks, v3, months & seasons

continuous calendar 2025 "months" (new) 

with AI illustrations in impressionist style (new)
US National Parks, 
with national holidays, 
and moon phases (new)
printable PDFs

English, with US official holidays,
and moon phases (EST), Letter size

English, Europe, A4 size

Serbian, latinica, A4 size

Serbian, ћирилица, holidays, moon phases, A4 size



"seasons" (winter, spring, summer, autumn/fall)









Posted by at No comments:

Postgres INSERT, 2x faster

using different SQL command can speed up insert 2x with Postgres DB

Boosting Postgres INSERT Performance by 50% With UNNEST

Benchmarking PostgreSQL Batch Ingest

INSERT INTO sensors (ts, sensorid, value) 
  SELECT * 
  FROM unnest(
    $1::timestamptz[], 
    $2::text[], 
    $3::float8[]
)
instead
INSERT INTO sensors (sensorid, ts, value)
VALUES 
  ($1, $2, $3), 
  ($4, $5, $6), 
   ..., 
  ($2998, $2999, $3000);
Posted by at No comments:
Subscribe to: Posts (Atom)