Jovani Pink
@jovanipink.bsky.social
Building scalable, compliant data platforms on Google Cloud that power analytics, machine learning, decision intelligence, and AI agents - driving measurable business growth.
Reposted by Jovani Pink
I thoroughly recommend reading all of Cory Doctorow's recent speech on AI skepticism, it's crammed with new arguments and interesting new ways of thinking about these problems pluralistic.net/2025/12/05/p...
Pluralistic: The Reverse-Centaur’s Guide to Criticizing AI (05 Dec 2025) – Pluralistic: Daily links from Cory Doctorow
pluralistic.net
December 7, 2025 at 10:21 PM
I thoroughly recommend reading all of Cory Doctorow's recent speech on AI skepticism, it's crammed with new arguments and interesting new ways of thinking about these problems pluralistic.net/2025/12/05/p...
Happy Thanksgiving!
November 27, 2025 at 2:12 PM
Happy Thanksgiving!
Life is a race. Because life is complicated and challenging, see, your heart constantly feels like it's breaking. The ONLY 🤣 time I've ever really run very long distances, I feel like my heart is just going to explode… that's life!
November 21, 2025 at 3:45 PM
Life is a race. Because life is complicated and challenging, see, your heart constantly feels like it's breaking. The ONLY 🤣 time I've ever really run very long distances, I feel like my heart is just going to explode… that's life!
Gemini 3!!! 🔥
Don't believe the hype that progress is slowing down. #ai #gcp #google
blog.google/products/gem...
Don't believe the hype that progress is slowing down. #ai #gcp #google
blog.google/products/gem...
A new era of intelligence with Gemini 3
Today we’re releasing Gemini 3 – our most intelligent model that helps you bring any idea to life.
blog.google
November 18, 2025 at 4:36 PM
Gemini 3!!! 🔥
Don't believe the hype that progress is slowing down. #ai #gcp #google
blog.google/products/gem...
Don't believe the hype that progress is slowing down. #ai #gcp #google
blog.google/products/gem...
Reposted by Jovani Pink
One of my core engineering beliefs is that if you are building something you should always be holistically looking at it
vickiboykis.com/2025/09/09/w...
vickiboykis.com/2025/09/09/w...
Walking around the app
You gotta check stuff out
vickiboykis.com
September 10, 2025 at 1:26 PM
One of my core engineering beliefs is that if you are building something you should always be holistically looking at it
vickiboykis.com/2025/09/09/w...
vickiboykis.com/2025/09/09/w...
Reposted by Jovani Pink
Reposted by Jovani Pink
Machine learning is built upon three pillars: linear algebra, calculus, and probability theory.
Here’s the full roadmap for you: thepalindrome.org/p/the-roadm...
Here’s the full roadmap for you: thepalindrome.org/p/the-roadm...
The Roadmap of Mathematics for Machine Learning
A complete guide to linear algebra, calculus, and probability theory
thepalindrome.org
August 24, 2025 at 5:57 PM
Machine learning is built upon three pillars: linear algebra, calculus, and probability theory.
Here’s the full roadmap for you: thepalindrome.org/p/the-roadm...
Here’s the full roadmap for you: thepalindrome.org/p/the-roadm...
Reposted by Jovani Pink
TIL you can run an eval suite against OpenAI's gpt-oss-20b open weights model running in LM Studio with the following uv one-liner. Here's my full TIL: https://til.simonwillison.net/llms/gpt-oss-evals
![mkdir /tmp/aime25_openai
OPENAI_API_KEY=x \
uv run --python 3.13 --with 'gpt-oss[eval]' \
python -m gpt_oss.evals \
--base-url http://localhost:1234/v1 \
--eval aime25 \
--sampler chat_completions \
--model openai/gpt-oss-20b \
--reasoning-effort low \
--n-threads 2](https://cdn.bsky.app/img/feed_thumbnail/plain/did:plc:mro7axagquvjt63foaqzddjx/bafkreiennb5rxhefjf7rmkonbvcu5oq2tppl2doyly34abnlkmmzdvnda4@jpeg)
August 17, 2025 at 3:49 AM
TIL you can run an eval suite against OpenAI's gpt-oss-20b open weights model running in LM Studio with the following uv one-liner. Here's my full TIL: https://til.simonwillison.net/llms/gpt-oss-evals
Treat each subagent as a pure function; the same input leads to the same statistical output; no shared memory or chat history exists. Parallelize safely, and agents don’t step on each other. Predictable behavior; test each agent in isolation. Cache via hash; keep state with the orchestrator. #ai
August 17, 2025 at 12:25 AM
Treat each subagent as a pure function; the same input leads to the same statistical output; no shared memory or chat history exists. Parallelize safely, and agents don’t step on each other. Predictable behavior; test each agent in isolation. Cache via hash; keep state with the orchestrator. #ai
I’ve been experimenting with running MCP servers locally, using tools like Claude Desktop, VS Code, and the mcphost library. To wire my own tools and data sources directly into LLM workflows. It's been working solidly, but I’m curious how others are approaching their MCP setup? #ai #MCP #AgenticAI
August 13, 2025 at 9:02 PM
I’ve been experimenting with running MCP servers locally, using tools like Claude Desktop, VS Code, and the mcphost library. To wire my own tools and data sources directly into LLM workflows. It's been working solidly, but I’m curious how others are approaching their MCP setup? #ai #MCP #AgenticAI
Reposted by Jovani Pink
OpenAI released their long-promised open weight models today under clean Apache 2 licenses and with benchmarks that put them shockingly close to o3-mini and o4-mini
I've run the smaller (20B) model on my Mac and it's very impressive, despite only using ~15GB of RAM simonwillison.net/2025/Aug/5/g...
I've run the smaller (20B) model on my Mac and it's very impressive, despite only using ~15GB of RAM simonwillison.net/2025/Aug/5/g...
OpenAI’s new open weight (Apache 2) models are really good
The long promised OpenAI open weight models are here, and they are very impressive. They’re available under proper open source licenses—Apache 2.0—and come in two sizes, 120B and 20B. OpenAI’s …
simonwillison.net
August 5, 2025 at 8:39 PM
OpenAI released their long-promised open weight models today under clean Apache 2 licenses and with benchmarks that put them shockingly close to o3-mini and o4-mini
I've run the smaller (20B) model on my Mac and it's very impressive, despite only using ~15GB of RAM simonwillison.net/2025/Aug/5/g...
I've run the smaller (20B) model on my Mac and it's very impressive, despite only using ~15GB of RAM simonwillison.net/2025/Aug/5/g...
Reposted by Jovani Pink
Today is my birthday (turning 29), which is the only time all year that I ask for a repost 🌿
I hope you get lost in fantasy worlds as much as I do 🔮
I hope you get lost in fantasy worlds as much as I do 🔮
July 31, 2025 at 12:12 PM
Today is my birthday (turning 29), which is the only time all year that I ask for a repost 🌿
I hope you get lost in fantasy worlds as much as I do 🔮
I hope you get lost in fantasy worlds as much as I do 🔮
Reposted by Jovani Pink
Just found out that George Orwell's Nineteen Eighty-Four, written around 1948, describes a "versificator" that composes the words for songs "entirely by mechanicsl means" simonwillison.net/2025/Jul/6/n...
A quote from Nineteen Eighty-Four
There was a whole chain of separate departments dealing with proletarian literature, music, drama, and entertainment generally. Here were produced rubbishy newspapers containing almost nothing except ...
simonwillison.net
July 6, 2025 at 9:44 AM
Just found out that George Orwell's Nineteen Eighty-Four, written around 1948, describes a "versificator" that composes the words for songs "entirely by mechanicsl means" simonwillison.net/2025/Jul/6/n...
Reposted by Jovani Pink
What’s the rationale for the name “DuckLake?”
As I understand it, you could one day have other compute engines like polars / Spark / some compute engine process data and use a Ducklake catalog on eg Postgres to manage the metadata (and parquet for the data).
What am I missing?
#duckdb #dataBS
As I understand it, you could one day have other compute engines like polars / Spark / some compute engine process data and use a Ducklake catalog on eg Postgres to manage the metadata (and parquet for the data).
What am I missing?
#duckdb #dataBS
June 7, 2025 at 6:15 PM
Reposted by Jovani Pink
From @nytopinion.nytimes.com
“Too many Christians are transforming Christianity into a vertical faith, one that focuses on your personal relationship with God at the expense of the horizontal relationship you have with your neighbors,” writes our columnist David French.
“Too many Christians are transforming Christianity into a vertical faith, one that focuses on your personal relationship with God at the expense of the horizontal relationship you have with your neighbors,” writes our columnist David French.
Opinion | Selfishness Is Not a Virtue
Evangelical Christianity has taken a wrong turn.
www.nytimes.com
June 5, 2025 at 9:14 PM
From @nytopinion.nytimes.com
“Too many Christians are transforming Christianity into a vertical faith, one that focuses on your personal relationship with God at the expense of the horizontal relationship you have with your neighbors,” writes our columnist David French.
“Too many Christians are transforming Christianity into a vertical faith, one that focuses on your personal relationship with God at the expense of the horizontal relationship you have with your neighbors,” writes our columnist David French.
Reposted by Jovani Pink
You can follow the original tutorial here: www.hackingwithswift.com/plus/live-st...
Here's the open-source project: github.com/twostraws/Sh...
There's also a channel for development discussion on the Hacking with Swift Slack workspace: www.hackingwithswift.com/slack
Here's the open-source project: github.com/twostraws/Sh...
There's also a channel for development discussion on the Hacking with Swift Slack workspace: www.hackingwithswift.com/slack
ShipShape – Hacking with Swift+
We're going to build an app that lets developers read data from App Store Connect – the apps they have, customer reviews, and more. This is going to require some careful networking, but there's lots t...
www.hackingwithswift.com
May 20, 2025 at 5:43 PM
You can follow the original tutorial here: www.hackingwithswift.com/plus/live-st...
Here's the open-source project: github.com/twostraws/Sh...
There's also a channel for development discussion on the Hacking with Swift Slack workspace: www.hackingwithswift.com/slack
Here's the open-source project: github.com/twostraws/Sh...
There's also a channel for development discussion on the Hacking with Swift Slack workspace: www.hackingwithswift.com/slack
Reposted by Jovani Pink
May 16, 2025 at 7:02 PM
Reposted by Jovani Pink
SCOOP from @tinanguyen.bsky.social: Elon's copyright coup attempt blew up after furious MAGA backlash to having AI companies train on content without paying for it www.theverge.com/politics/666...
Elon Musk’s apparent power play at the Copyright Office completely backfired
Trump fired the head of the Copyright Office over the weekend.
www.theverge.com
May 13, 2025 at 8:50 PM
SCOOP from @tinanguyen.bsky.social: Elon's copyright coup attempt blew up after furious MAGA backlash to having AI companies train on content without paying for it www.theverge.com/politics/666...
Reposted by Jovani Pink
Just watched a short-form video of someone explaining the relation of square roots to areas using sidewalk chalk while their kids yell 'Daddy!' to get their attention & them responding 'just one second girls, I'm drawing a perfect square!'
I had an honest moment of 'wait, is this literally me?'
I had an honest moment of 'wait, is this literally me?'
May 13, 2025 at 1:39 PM
Just watched a short-form video of someone explaining the relation of square roots to areas using sidewalk chalk while their kids yell 'Daddy!' to get their attention & them responding 'just one second girls, I'm drawing a perfect square!'
I had an honest moment of 'wait, is this literally me?'
I had an honest moment of 'wait, is this literally me?'
Reposted by Jovani Pink
OpenAI released o3 and o4-mini today with a strong emphasis on tool use. I've upgraded my llm-openai plugin to support the two new models. simonwillison.net/2025/Apr/16/...
Introducing OpenAI o3 and o4-mini
OpenAI are *really* emphasizing tool use with these: > For the first time, our reasoning models can agentically use and combine every tool within ChatGPT—this includes searching the web, analyzing …
simonwillison.net
April 16, 2025 at 5:48 PM
OpenAI released o3 and o4-mini today with a strong emphasis on tool use. I've upgraded my llm-openai plugin to support the two new models. simonwillison.net/2025/Apr/16/...
Reposted by Jovani Pink
Model Context Protocol has prompt injection security problems
simonwillison.net/2025/Apr/9/m...
simonwillison.net/2025/Apr/9/m...
Model Context Protocol has prompt injection security problems
As more people start hacking around with implementations of MCP (the Model Context Protocol, a new standard for making tools available to LLM-powered systems) the security implications of tools built ...
simonwillison.net
April 9, 2025 at 1:01 PM
Model Context Protocol has prompt injection security problems
simonwillison.net/2025/Apr/9/m...
simonwillison.net/2025/Apr/9/m...
Reposted by Jovani Pink
Wrote up my first impressions of the new GPT-4.5 - it's quite slow, VERY expensive and doesn't appear to be a notable leap forward from GPT-4o or o3-mini simonwillison.net/2025/Feb/27/...
Introducing GPT-4.5
GPT-4.5 is out today as a "research preview" - it's available to OpenAI Pro ($200/month) customers but also to developers with an API key. OpenAI also published [a GPT-4.5 system …
simonwillison.net
February 27, 2025 at 9:26 PM
Wrote up my first impressions of the new GPT-4.5 - it's quite slow, VERY expensive and doesn't appear to be a notable leap forward from GPT-4o or o3-mini simonwillison.net/2025/Feb/27/...