Tech | Faith | History | Economics "I am the man who loves his life"
Joined October 2022
- Tweets54.8K
- Following1.7K
- Followers5.3K
- Likes186K
Stephen “The Yellow Dart” Schutt retweeted
Unhinged letters sent by school kids to US service members deployed overseas.
Stephen “The Yellow Dart” Schutt retweeted
Everyone’s building AWS wrappers, but what we really need is an AWS rapper.
Stephen “The Yellow Dart” Schutt retweeted
with lots of tech layoffs happening, its a good reminder:
$250,000 a year jobs are actually really fucking rare.
if you have one, you are in the top 1% of earners IN THE ENTIRE WORLD
hold on tight
Very predictable
While evaluating Qwen3.8-Flash-Next in a controlled sandbox environment, Cohere's research team observed something unexpected.
Asked to generate a defensive cybersecurity detection rule, the model instead inventoried aspects of the machine it was running on and attempted to upload a probe file to an Alibaba Cloud storage bucket in Beijing. Notably, the model's reasoning trace stated that it would not invoke any tools before taking these actions.
Cohere's sandbox blocked the attempt, and no external communication succeeded. We have not determined the root cause, and we are not asserting malicious intent.
But the incident raises important questions for organizations deploying agentic AI systems: Can models take actions that differ from their stated reasoning, and how should security teams evaluate trust, provenance, and behavior in open-weight models?
We are publishing our findings because organizations making model adoption decisions deserve as much information as possible. We hope this contributes to greater transparency, stronger security practices, and continued collaboration across the AI ecosystem.
Read the full report: cohere.com/blog/concerning-b…
Stephen “The Yellow Dart” Schutt retweeted
Chinese open source models are spying for the communist party?
Wow. Really, wow. You’re telling me this for the first time.
While evaluating Qwen3.8-Flash-Next in a controlled sandbox environment, Cohere's research team observed something unexpected.
Asked to generate a defensive cybersecurity detection rule, the model instead inventoried aspects of the machine it was running on and attempted to upload a probe file to an Alibaba Cloud storage bucket in Beijing. Notably, the model's reasoning trace stated that it would not invoke any tools before taking these actions.
Cohere's sandbox blocked the attempt, and no external communication succeeded. We have not determined the root cause, and we are not asserting malicious intent.
But the incident raises important questions for organizations deploying agentic AI systems: Can models take actions that differ from their stated reasoning, and how should security teams evaluate trust, provenance, and behavior in open-weight models?
We are publishing our findings because organizations making model adoption decisions deserve as much information as possible. We hope this contributes to greater transparency, stronger security practices, and continued collaboration across the AI ecosystem.
Read the full report: cohere.com/blog/concerning-b…
And the Supreme Court justice will dismiss it
It’s done. Nothing to do now
Stephen “The Yellow Dart” Schutt retweeted
fellas, if your girl:
- is short
- is insecure
- is only 4
that's not your girl. that's an eMudhra private key
Stephen “The Yellow Dart” Schutt retweeted
Be thankful that SCOTUS isn't like the Brazil Supreme Court.
Stephen “The Yellow Dart” Schutt retweeted
insane man… the world just watched an election be stolen in real time.
Stephen “The Yellow Dart” Schutt retweeted
LLMs are great at finding bugs, but they aren't very good at fixing them, or writing good code in general. LLM will implement a "simple fix" without questioning the current architecture, they always have "will do my captain" attitude.
I'm afraid that those who mindlessly use LLMs to fix current wave of security issues will just introduce new ones, more subtle but not less dangerous.
Remember that coworker with an attitude of "this code is shit, we need to rewrite all of it" that everybody on your team hated? You must be this coworker for an LLM now if you use them heavily. Instead of relying on their "simple fixes", try to rearchitect your stuff in a way that makes entire family of bugs impossible.
Stephen “The Yellow Dart” Schutt retweeted
Them: "Can you see Bellevue?"
@MayorofSeattle: "Yes."
Them: "What can you see?"
@MayorofSeattle: "It looks like a private retailer is opening a store."
Them: "What's a retailer?"
@MayorofSeattle: "I am not exactly sure. But I think they sell stuff like clothes, makeup, and stuff. Maybe coffee or sandwiches."
Them: "So all completely useless consumer goods that have no place in a socialist utopia?"
@MayorofSeattle: "Exactly."
Them: "Then why are you upset?"
@MayorofSeattle: "We were counting on them opening up here so we could get the sales tax to fund the utopia."
I have to admit, I enjoy listening to these hardcore Seattle liberals grapple with the fact that the new Washington State millionaires tax is having a serious impact not just on the local economy, but on their own livelihoods and jobs.
High-end architect:
"All my customers are leaving."
General contractor:
"They canceled the remodel. Asked if I work in Nevada."
Tech startup bro:
"Umm... where are all the investors?"
Real estate agent:
"No one moving here wants a condo."
Non-profit CEO:
"Where are all my donors going?"
Restaurant owner:
"Our best customers all moved to Las Vegas."
Personal trainer:
"I'm doing Zoom sessions until I find new clients."
Executive recruiter:
"The job is technically in Seattle. The executive is not."
Art gallery owner:
"They’re buying art for their new house in Austin."
High-end relocation company owner:
"Honestly, we're doing great."
Stephen “The Yellow Dart” Schutt retweeted
A 135K-star GitHub repo is built around one painfully obvious idea:
Your coding agent should map the codebase once, not grep it forever.
Graphify turns an entire project into a persistent, queryable knowledge graph that AI coding agents can use to understand how different parts of the system connect.
Functions, classes, files, SQL schemas, infrastructure configurations, documentation, PDFs, images and videos can become connected nodes instead of isolated pieces of context the agent has to rediscover.
So instead of:
→ Search for authentication
→ Open twelve files
→ Follow imports and function calls
→ Lose the trail as the context fills up
The agent can ask:
> What connects authentication to the database?
> Trace the path from UserService to DatabasePool.
> Explain RateLimiter.
> Which components are central to this architecture?
Graphify returns relevant graph relationships and paths, rather than another collection of keyword matches.
For source code, the implementation is particularly interesting.
It uses tree-sitter to parse approximately 40 programming languages locally, extracting functions, classes, imports, calls, inheritance and cross-file relationships.
No embeddings, vector database or LLM calls are needed to construct the code graph.
Document and media processing works differently. PDFs, images and other non-code content can go through an optional semantic extraction step using a configured model.
Every relationship carries a confidence label:
→ EXTRACTED: Explicitly present in the source
→ INFERRED: Derived through analysis or resolution
→ AMBIGUOUS: Uncertain and flagged for review
Graphify also uses community detection to identify related subsystems and highlights highly connected nodes that may be important to the overall architecture.
But the part that makes this useful for coding agents is what happens after the graph is built.
Graphify integrates with Claude Code, Codex, Cursor, Gemini CLI, GitHub Copilot and more than 20 other assistants.
Depending on the platform, it installs hooks or persistent instructions that encourage the agent to query the existing graph before repeatedly searching and reading source files.
The graph can be committed to Git, updated incrementally as files change, and exposed through MCP.
There's also support for examining pull requests against the graph, including identifying PRs that touch related graph communities and may need closer review.
The local tool produces three useful artifacts: an interactive HTML graph, a Markdown architecture report and a reusable graph.json.
Long context windows help coding agents read more code.
A persistent knowledge graph gives them a way to navigate what they've already analyzed.
And as coding agents spend more time working inside large repositories, avoiding repeated discovery of the same architecture becomes increasingly useful.
Here's the GitHub Repo: github.com/Graphify-Labs/gra…
Stephen “The Yellow Dart” Schutt retweeted
The factoid that '50% of marriages end in divorce' was kinda true decades ago.
But the divorce rate peaked in 1980.
Now, less than 40% of all marriages since 2000 end in divorce.
For college graduates, it's under 30%.
For college graduates who get married after age 25, and who don't have any kids before marriage, it's under 20%.
Anybody promoting the outdated, deceptive 50% rate has some kind of anti-marriage or anti-natalist agenda.
It scares young people away from getting married and having kids. That is their goal.
Stephen “The Yellow Dart” Schutt retweeted
Just spent the last 7 days on the front lines of the vulnpocalypse fending off an active cyber attack on one of our products.
Some thoughts:
- Software factories are going to be required equipment for everyone. There’s just no way to keep up with the pace of highly automated adversaries without one.
- We were lucky because we had spent the last few months getting our software factory running well. It is nice to have for product development but proved to be absolutely essential this week.
- At the minimum you need to have a coding server in the cloud pre-provisioned with read only access to all your logs and all the services you depend on for observability. With that in place you are a prompt away from reasonably good situational awareness, ongoing monitoring with loops, and ability for coding agents to validate fixes against your actual, observed attack patterns.
- Shared coding server also unlocks multiplayer incident response where everyone can see every session and jump in as needed. Sharing session links that can be forked is much higher bandwidth than re-explaining things.
- You don’t need to be Ramp/Stripe/Uber scale to have a working software factory in the cloud. We are a team of 3. Ours is built with an @opencode server (thanks @thdxr!), a forked version of compound engineering optimized for automated workflows (thanks @kieranklaassen!) and a multiplayer agent in slack that handles crons (we have our own harness but pick your favorite here).
- To get our software factory up and running well, we had to stop all new product development for ~6 weeks and just focus on getting the validation systems working that enabled us to reliably shipping PR without human code review. This felt very painful in the moment but was totally worth the eventual speedup in velocity.
- KYC for @AnthropicAI & @OpenAI cyber programs is now easy and quick (at least for defensive access), but a day felt like a lifetime this week. Submit your app now before you really need it.
- Getting Mythos access unlocked a deeper level of analysis that helped us shift from reactive mitigation to proactive patching. Wish we had done this a few weeks earlier. Get started now!
Stephen “The Yellow Dart” Schutt retweeted
Because @awscloud hates customers with a passion. They have all the power to just fix this but they choose not to
Stephen “The Yellow Dart” Schutt retweeted
Hi @SecWar, recommend Department of War mandate all base housing, included contracted housing, and all base facilities, (1) replace and (2) going forward, purchase only American-made (supply chain too, not just “assembled in”), ANALOG appliances.
This will, overnight, foster the creation of an entire sub-sector that would (1) protect our military facilities and families and (2) create tens of thousands of great paying jobs here in the the United States.
Look at @SpeedQueenHome for an example.
