Software Developer, Book Author, JavaFX fan, and love AI. Most importantly a husband & proud father. My opinions are my own.
The 'Dena' in Mary-land, USA
Joined December 2008
- Tweets18.8K
- Following2.3K
- Followers2.7K
- Likes30.1K
Carl Dea ⒻⓍ🕊 retweeted
New video on Goose: youtube.com/watch?v=kHDsfGvM…
Carl Dea ⒻⓍ🕊 retweeted
Every Mac on macOS 27 now has a free LLM built into the Terminal.
No account. No API key. No internet. No token bill.
Setup (1 minute):
1. sudo fm license (enter your Mac password, one time)
2. fm available (checks the model is downloaded)
3. fm respond “your question”
Works on any Apple Silicon Mac, M1 included.
- Pipe files: cat notes.txt | fm respond -i “summarize this”
- Get JSON back: build a schema with fm schema object, pass it with --schema
- Local API: fm serve gives you an OpenAI-compatible endpoint on localhost
- Resume chats: sessions save to ~/.fm/sessions/
The limit is 4,096 tokens per session, and your prompt, instructions, and the answer all share it.
So skip code and math. Use it for summarizing, extracting, classifying, and rewriting.
Free, private, offline, and already on your Mac.
Carl Dea ⒻⓍ🕊 retweeted
@OpenJDK The list of JDK 28 looks very intresting JEP 543 Structured Concurrency is now finalized (Status: Integrated)
openjdk.org/projects/jdk/28/
Carl Dea ⒻⓍ🕊 retweeted
Fetching JSON from an API and mapping it to records in plain Java.
No Jackson. No Gson. No dependencies.
Just HttpClient and the new JSON API coming in JDK 28 (JEP 540). 👀
github.com/danvega/native-js…
Carl Dea ⒻⓍ🕊 retweeted
Introducing @Pyronautfw — a high-performance framework for Python, built with Micronaut, GraalVM, and GraalPy!🚀
By @graemerocher and @thomaswue at #Devoxx: youtube.com/watch?v=zo6L7bxn…
Carl Dea ⒻⓍ🕊 retweeted
Start in 1972. After the 1970 census, Maryland drew congressional maps that respected geographic and political boundaries. The result was a delegation of 4 Republicans and 4 Democrats, a genuinely competitive split reflecting Maryland’s actual political makeup. Fair maps produce fair elections.
Carl Dea ⒻⓍ🕊 retweeted
WebMCP by Microsoft and Google is HERE and I'm surprised more people aren't talking about it.
Why does it matter?
BILLIONS of dollars are about to move through agents, and the internet isn't built for them yet.
1. SEO was about Google understanding your page.
2. AEO was about AI citing you.
3. WebMCP is about the agent actually finishing the job (buying, researching, requesting quote etc).
WebMCP is basically websites with agent buttons.
Instead of an agent scanning a page like a human, screenshotting and guessing where to click, the site just tells it "here's how to search, here's how to book, here's how to buy."
2 cash flowing businesses you could start today using WebMCP:
1. A WebMCP conversion agency
Make boring business websites agent-ready. Law firms, HVAC, med spas, dentists. Build them the first tools (request a quote, book a consult), sell the setup for $2k, then charge a few hundred a month to monitor and improve it.
2. An agent mystery shopper.
Test whether agents can actually complete the important journeys on someone's site, buying the hoodie, booking the consult, filing the claim. Hand them a report on where the agent got stuck, missing tools, bad descriptions, lost conversions. Charge monthly, then turn the repeated fixes into software.
Full episode on @startupideaspod with @hot_town below
youtu.be/EoNH3Tn8wYE?si=vqaN…
You'll learn what WebMCP really is, see examples of it IN ACTION, and hear more about 2 startup ideas using WebMCP you can start today.
WebMCP is pretty cool.
Watch.
Carl Dea ⒻⓍ🕊 retweeted
The AI containment debate may already be over.
@friedberg argument is that powerful AI is no longer controlled by a handful of American companies running enormous data centers.
Open weight models are becoming capable and efficient enough to download, modify, and run on consumer hardware, making the technology nearly impossible to contain once released.
The recent launch cycle shows how quickly this is happening.
DeepSeek released a faster and cheaper model, Alibaba published downloadable weights for Qwen Image 2.1, and PrismML compressed a 27 billion parameter Qwen model into just 5.9 gigabytes.
PrismML says its compressed model retained 98.2% of the original model’s benchmark performance while becoming small enough to run on a laptop or consumer GPU.
At the same time, closed model companies continued releasing more capable and affordable systems.
Anthropic’s Opus 5.5 reportedly matched its more expensive flagship while costing approximately 40% less to run than its predecessor.
Stanford found that the cost of running a model with GPT 3.5 level performance fell more than 280 fold in roughly 18 months.
It also found that the performance gap between leading open-weight and closed models narrowed from 8% to 1.7% on some benchmarks in one year.
That is the real meaning of Friedberg’s argument.
Governments can regulate large data centers, chip exports, and commercial AI providers but they cannot easily remove models that have already spread across the internet and onto personal computers.
The debate must therefore shift from whether powerful AI should exist to how it should be used and who becomes responsible when it causes harm.
Carl Dea ⒻⓍ🕊 retweeted
a Stanford professor used JEV to build a data science system that analyzes 40.000.000.000 data points every 15 min
his system can process massive datasets and automatically decide which results deserve deeper analysis.
the professor spent more than a year developing the system with his research team.
the first version relied heavily on LLMs to interpret every result.
it worked, but analyzing millions of records created unnecessary latency and huge inference costs.
then they redesigned the decision layer around Jev.
instead of generating explanations for every data point, Jev evaluates the results and decides what the system should investigate next.
I found a breakdown of the pipeline they used and the way Jev filters the data is surprisingly clever.
I’ll show you how it works next so stay tuned if you want to see what Jev Engineering looks like inside a real data science system.
would you let an AI decide which patterns in your data are worth investigating?
Carl Dea ⒻⓍ🕊 retweeted
Caveman Issues Dire Warning That The Wheel Is Going To Eliminate Countless Jobs buff.ly/u9nCQoL
Carl Dea ⒻⓍ🕊 retweeted
Day 12 building the Figma for hardware until somebody invests in it.
This week I’m focusing entirely on jewelry.
Technology is getting closer and closer to our bodies, and I think the objects around it will matter just as much as the technology itself. A necklace, ring or charm can carry something digital without asking for your attention all the time.
So I want to explore the space between hardware, jewelry and everyday objects: connectors, chains, rings, charms, carriers and ways of turning technology into something you actually want to wear.
The system I’m building will eventually go beyond devices. I want it to be a way to design objects in general.
Carl Dea ⒻⓍ🕊 retweeted
I’d like to announce two projects I’m working on:
First, I am launching a non-profit with the aim of gathering Christians in tech to help build “software for the kingdom”. There isn’t much on the site yet, but to stay in the loop, drop your email --> awakenusa.com.
Second, I have taken on the role as president for a new TPUSA Communities San Francisco Bay Area chapter. I’ll be focused on building stronger local communities through social and civic engagement, volunteer opportunities, and events, all united behind the shared values of faith, family, and freedom. More on this --> tpusasfbay.com.
What led me here?
Today marks one year since Charlie Kirk was assassinated. His death had a profound impact on me. I know many disagree with his views, but having listened to him for years on campuses, on air, and in churches, there’s no denying his unwavering commitment to dialogue, truth, and freedom.
He gave everyone a mic, especially those who disagreed with him, listened intently, and treated them with respect and decency. And in his last few years his faith was front and center, boldly proclaiming the gospel. I can see the fruits of this in our own church, especially with the younger generation.
I’ve been in tech for almost 30 years, much of that running my own businesses. But over the last several years, something has shifted. AI brings hope, but also separation. Truth and freedom have been inverted, giving rise to deception, anger, and loneliness. As technologists, we are builders and optimists at heart, but it’s hard to ignore the negative consequences.
What world will my three kids grow up in? Can technology be used for good or are we just building for profit? Will technology unite us or divide us? Will it be used to increase agency or control us?
This is where faith and tech must find a deeper union.
And this is what led me to create Awaken and team up with TPUSA.
As the sun rises on 9/10, and tomorrow on 9/11, and soon 10/7, so many people remain divided, confused, and lonely. As technologists and entrepreneurs, let's use our gifts and talents to inspire dialogue over division, to seek truth over deception, and to stand for freedom above control!
Ultimately let’s “be not overcome of evil, but overcome evil with good”, and to “Let your light so shine before men, that they may see your good works, and glorify your Father which is in heaven.”
Here I am Lord. Send me.
Carl Dea ⒻⓍ🕊 retweeted
For years, Maryland was considered the poster child for extreme partisan gerrymandering in the nation. Our congressional districts were so unfair, twisted, and distorted that they became a punchline on every list of the worst-drawn maps in America.
We appointed an independent, non-partisan redistricting commission that held hearings all across the state and heard the testimony of thousands of Marylanders. The legislature's extreme partisan maps were thrown out by the courts, and the legislature was ordered to redraw fairer maps. These are the ones we currently use today.
Our fair maps took eight years to win in Maryland, and the effort received bipartisan praise across the country.
Now, to ensure that their partisan, illegal maps can never be thrown out again, the monopoly in Annapolis is attempting to rewrite our state Constitution to clear the way for the same kind of maps that the courts already said violated the law.
Let me be clear, the way I've been clear about this for decades: partisan gerrymandering is a cancer on our democracy, and it's wrong no matter who does it.
Marylanders, regardless of party, should say NO to extreme partisan gerrymandering. NO to politics as usual. NO to being the worst in America. We are better than this.
wbaltv.com/article/redistric…
Carl Dea ⒻⓍ🕊 retweeted
META'S CHIEF SCIENTIFIC ADVISOR, YANN LECUN, SAID, "CHATGPT, CLAUDE, AND ALL CURRENT LANGUAGE MODELS ARE A DEAD END, AND INVESTORS ARE WASTING TRILLIONS OF DOLLARS."
This scandalous statement was made by a Turing Award winner and a man considered one of the "godfathers" of modern AI, which has the entire IT industry in stitches.
While IT giants are buying up hundreds of thousands of NVIDIA graphics cards and promising "superintelligence by next year," LeCun is blunt: attempting to create strong artificial intelligence (AGI) based on LLM is a fundamental misconception for the entire industry.
Here are four compelling arguments from an AI legend about why the chatbot era has hit a wall:
• Lack of a model of the real world: A 4-month-old baby or a cat understands physics, gravity, and space better than the most powerful neural network. Training AI solely on text is like trying to learn to drive by forcing a person to read the instructions on the couch.
• The Illusion of Reason Instead of Thinking: Current LLMs are simply mathematical next-word predictors (pattern matching). They lack logic, true memory, and planning ability. The models produce plausible text, but have no understanding of its semantic essence.
• The Scaling Wall: Increasing data centers and budgets by 10x yields only a symbolic increase in quality. The industry is trying to squeeze the maximum out of a technology that has reached its architectural limits.
• The Shift to "World Models": True AGI will emerge not from text chats, but from fundamentally new architectures-"world models"-capable of perceiving physical reality and cause-and-effect relationships, just as the human brain does.
Carl Dea ⒻⓍ🕊 retweeted
Jeff Hinton dismissed gradient descent in the 1970s because he thought models would get stuck. He won the Nobel Prize in 2024 partly for being wrong about that.
Every time ChatGPT gets better after a training run, it's because 1.2 billion parameters fell through a wormhole in a loss landscape - not because they crawled downhill one step at a time. This video visualizes exactly what that looks like inside Llama 3.2.
20 minutes. Bookmark & watch today. The misconception that almost stopped AI is more interesting than the solution.
Carl Dea ⒻⓍ🕊 retweeted
The open-source @DevoxxGenie @JetBrains Coding Agent plugin has just surpassed 104,000 active users and over 80,000 Marketplace downloads! 🚀🥳
A huge thank-you to everyone using it and contributing feedback 🙏❤️
See also genie.devoxx.com