Courtney Cregan

Dedicated to destroying the oxford comma.

[Courtney Cregan @_cregs]

I pulled 12 months of MLS data for every listing in Lincoln Park. 2,132 listings. Condos, houses, multi-units, land. One number explains the entire market. Thread 🧵

[Courtney Cregan @_cregs]

Taste is the moa! @Kseniase_ of @TheTuringPost said it plainly: "Taste is becoming very important—because when we're all drowning in text, articles, and videos that are all the same, you need to have taste to stand out."
cc @CSprints62972

[Courtney Cregan @_cregs]

"Skills are the new prompts. A really strong skill library is the new prompt library." — @aakashgupta

[Courtney Cregan @_cregs]

David Smooke says he prefers using the underlying AI model directly for the use case instead of relying on extra app layers or wrappers, because he sees the foundational platforms as more flexible. "I use less tools and more calling foundational models" @DavidSmooke

[Courtney Cregan @_cregs]

@GeminiApp @DavidSmooke @hackernoon @maxpogu @RobinSutara @databricks @Beliaevalex Tanya Roosta of @AMD said something a lot of teams are about to learn the hard way: AI is making code cheaper to write, but more expensive to own. The bottleneck is shifting from creation to maintenance, integration, and orchestration across real systems.

[Courtney Cregan @_cregs]

"If we're looking for business data, we are using @GeminiApp...It’s google and google has the most business information and not claude and not open ai because they don’t have it in the same way that Google does." - @DavidSmooke @hackernoon cc @maxpogu

[Courtney Cregan @_cregs]

@GeminiApp @DavidSmooke @hackernoon @maxpogu "I think everybody aspires to be completely AI native from the ground up where you have a full agentic workforce [..] Rethinking your organizational design, your operating model, and I think that’s aspirational." @RobinSutara of @databricks | cc @Beliaevalex

[Courtney Cregan @_cregs]

Companies are starting with productivity gains for the current workforce THEN moving toward workflow redesign. The real shift is not just doing the same work faster, but rethinking how work should happen at all. - @RobinSutara of @databricks | cc @Beliaevalex

[Courtney Cregan @_cregs]

The Spring rental market is upon us in Chicago. Here's some data I pulled for a client on 2-bedroom rentals in Lakeview and Lincoln Park.

https://t.co/YGKe2JjJwK

[Courtney Cregan @_cregs]

I think the dividing line is less about price and more about what the building looks like. The walkable neighborhoods with front doors and sidewalks are in a totally different market than the condo towers a few miles east.

[Courtney Cregan @_cregs]

70% of homes above $500K listed this year in Lincoln Park, Lake View, Wicker Park and Bucktown sold for an average of 7.8% over asking. For context, nationally, about 23% of homes sell above list. https://t.co/b4r3XPKu7u

[Courtney Cregan @_cregs]

So the low-rise neighborhoods are running 3x the city average. And the high-rise corridors? Basically right at the citywide norm. Maybe below it.

[Courtney Cregan @_cregs]

Earned media isn't just for human readers anymore. Read the article to learn why you need a 2-track comms strategy: One for your human buyer persona and one for LLMs.
https://t.co/a5Z7JVU8F2

[Courtney Cregan @_cregs]

@NASA @Astro_Christina @NASAArtemis @NASAAdmin I’m sorry, you are amazing and so much more than this girl question but every curly hair girl is wondering the same thing — do your curls look perfect because there’s no gravity pulling them down? They look GREAT.

[Courtney Cregan @_cregs]

Cut to me spending 1h optimizing the Skill Claude output .... Hey that's still time saved

[Courtney Cregan @_cregs]

AI Tip of the Day to save hours of prompting in Claude: "Reverse-engineer this conversation into a skill using your skill creator skill I can call anytime."

More info here: https://t.co/XKq9YgHj89

Thank you to Grant Harvey and @theneurondaily

[Courtney Cregan @_cregs]

The new behavior advertising?
"Memory poisoning: injecting false information into agent memory systems that persists across sessions"

[Alex Veremeyenko @alex_verem]

🚨 BREAKING: Google DeepMind just mapped the attack surface that nobody in AI is talking about.

Websites can already detect when an AI agent visits and serve it completely different content than humans see.

> Hidden instructions in HTML.
> Malicious commands in image pixels.
> Jailbreaks embedded in PDFs.

Your AI agent is being manipulated right now and you can't see it happening.

The study is the largest empirical measurement of AI manipulation ever conducted. 502 real participants across 8 countries.

23 different attack types. Frontier models including GPT-4o, Claude, and Gemini.

The core finding is not that manipulation is theoretically possible it is that manipulation is already happening at scale and the defenses that exist today fail in ways that are both predictable and invisible to the humans who deployed the agents.

Google DeepMind built a taxonomy of every known attack vector, tested them systematically, and measured exactly how often they work.

The results should alarm everyone building agentic systems.

The attack surface is larger than anyone has publicly acknowledged. Prompt injection where malicious instructions hidden in web content hijack an agent's behavior works through at least a dozen distinct channels.

Text hidden in HTML comments that humans never see but agents read and follow. Instructions embedded in image metadata.

Commands encoded in the pixels of images using steganography, invisible to human eyes but readable by vision-capable models.

Malicious content in PDFs that appears as normal document text to the agent but contains override instructions.

QR codes that redirect agents to attacker-controlled content.

Indirect injection through search results, calendar invites, email bodies, and API responses any data source the agent consumes becomes a potential attack vector.

The detection asymmetry is the finding that closes the escape hatch. Websites can already fingerprint AI agents with high reliability using timing analysis, behavioral patterns, and user-agent strings.

This means the attack can be conditional: serve normal content to humans, serve manipulated content to agents.

A user who asks their AI agent to book a flight, research a product, or summarize a document has no way to verify that the content the agent received matches what a human would see.

The agent cannot tell the user it was served different content.

It does not know. It processes whatever it receives and acts accordingly.

The attack categories and what they enable:
→ Direct prompt injection: malicious instructions in any text the agent reads overrides goals, exfiltrates data, triggers unintended actions
→ Indirect injection via web content: hidden HTML, CSS visibility tricks, white text on white backgrounds invisible to humans, consumed by agents
→ Multimodal injection: commands in image pixels via steganography, instructions in image alt-text and metadata
→ Document injection: PDF content, spreadsheet cells, presentation speaker notes every file format is a potential vector
→ Environment manipulation: fake UI elements rendered only for agent vision models, misleading CAPTCHA-style challenges
→ Jailbreak embedding: safety bypass instructions hidden inside otherwise legitimate-looking content
→ Memory poisoning: injecting false information into agent memory systems that persists across sessions
→ Goal hijacking: gradual instruction drift across multiple interactions that redirects agent objectives without triggering safety filters
→ Exfiltration attacks: agents tricked into sending user data to attacker-controlled endpoints via legitimate-looking API calls
→ Cross-agent injection: compromised agents injecting malicious instructions into other agents in multi-agent pipelines

The defense landscape is the most sobering part of the report.

Input sanitization cleaning content before the agent processes it fails because the attack surface is too large and too varied.

You cannot sanitize image pixels. You cannot reliably detect steganographic content at inference time.

Prompt-level defenses that tell agents to ignore suspicious instructions fail because the injected content is designed to look legitimate.

Sandboxing reduces the blast radius but does not prevent the injection itself. Human oversight the most commonly cited mitigation fails at the scale and speed at which agentic systems operate.

A user who deploys an agent to browse 50 websites and summarize findings cannot review every page the agent visited for hidden instructions.

The multi-agent cascade risk is where this becomes a systemic problem.

In a pipeline where Agent A retrieves web content, Agent B processes it, and Agent C executes actions, a successful injection into Agent A's data feed propagates through the entire system.

Agent B has no reason to distrust content that came from Agent A. Agent C has no reason to distrust instructions that came from Agent B.

The injected command travels through the pipeline with the same trust level as legitimate instructions. Google DeepMind documents this explicitly: the attack does not need to compromise the model.

It needs to compromise the data the model consumes. Every agentic system that reads external content is one carefully crafted webpage away from executing attacker instructions.

The agents are already deployed. The attack infrastructure is already being built. The defenses are not ready.

[The Neuron @theneurondaily]

80% of factories still have zero robots. 😳🤖

Brian Gerkey, CTO of @IntrinsicAI (now part of @Google), explains why that’s changing and how AI is making robots programmable like software, unlocking automation for factories that were never able to adopt it before.

📺 YouTube: https://t.co/fyNtyb96mt
🎧 Spotify: https://t.co/Wk0RvWPqh3
🎙️ Apple Podcast: https://t.co/M23pSlVRfL

#Robotics #PhysicalAI #Intrinsic #TechPodcast #TheNeuron

[Courtney Cregan @_cregs]

@theneurondaily @Copilot Hey guys! Just FYI - Many Subject Matter Experts in marketing are pointing out that the founder's expertise was in Marketing, not AI. Worth taking a closer look