jkoblelus.bsky.social
@jkoblelus.bsky.social
Pinned
My "Pieces of Fate" series of poetry books is now on sale through Kindle. Purchase them here: shorturl.at/SFnZK. Let me know what you think.
Amazon.com : james kobielus
shorturl.at
"Debt Has Entered the A.I. Boom" (shorturl.at/bRw4Z) JK--Yeah, but excessive debt is precisely the thing that could make all this AI mania go totally ka-BOOM!
shorturl.at
November 24, 2025 at 8:02 PM
"The state of AI in 2025: Agents, innovation, and transformation" (shorturl.at/X223w) JK--Interesting research from McKinsey.
shorturl.at
November 24, 2025 at 7:59 PM
"Beyond Standard LLMs" (shorturl.at/916LW) JK--Very good, very technical article about recent evolutionary radiation out from and beyond the breakthrough autoregressive transformer architecture.
Beyond Standard LLMs
Linear Attention Hybrids, Text Diffusion, Code World Models, and Small Recursive Transformers
shorturl.at
November 18, 2025 at 7:47 PM
"Special Edition: State of AI Ethics Report (SAIER) Volume 7" (shorturl.at/gqC2r) JK--OK. It'd be nice to have a State of Human Ethics Report to benchmark how more or less virtuous than us our technological contrivances are becoming.
Special Edition: State of AI Ethics Report (SAIER) Volume 7
AI at the Crossroads: A Practitioner's Guide To Community-Centered Solutions
shorturl.at
November 18, 2025 at 7:19 PM
"Introducing Aardvark: OpenAI’s agentic security researcher" (shorturl.at/sokbT) JK--"Aardvark ....uses LLM-powered reasoning and tool-use to understand code behavior and identify vulnerabilities."
Introducing Aardvark: OpenAI’s agentic security researcher
Now in private beta: an AI agent that thinks like a security researcher and scales to meet the demands of modern software.
shorturl.at
November 18, 2025 at 7:15 PM
"On AI Slop vs OSS Security" (shorturl.at/GFKoR) JK--"A "security researcher" who just pastes LLM-generated output into a vulnerability submission form neither knows enough about the actual codebase being examined nor understands the security implications well enough."
On AI Slop vs OSS Security
I have now spent almost a decade in the bug bounty industry, started out as a bug hunter (who initially used to submit reports with minimal impact, low-h...
shorturl.at
November 18, 2025 at 6:52 PM
"Meet Project Suncatcher, Google’s plan to put AI data centers in space" (shorturl.at/Zmgw8) JK--Just what we need. More potential space junk imperiling human efforts to escape the gravitational clutch of our own planet. Now it'll be data-center shrapnel in low earth orbit.
Meet Project Suncatcher, Google’s plan to put AI data centers in space
Google is already zapping TPUs with radiation to get ready.
shorturl.at
November 18, 2025 at 6:41 PM
New Kobielus LinkedIn thought-leadership post: "How Agentic AI Is Driving Evolution of the Enterprise Data Fabric" (tinyurl.com/2k4vweab)
How Agentic AI Is Driving Evolution of the Enterprise Data Fabric
Agentic AI applications are driving architectural innovations in practically every computing domain. In recent months, I’ve been noticing an uptick in discussions of how databases will evolve to suppo...
tinyurl.com
November 6, 2025 at 8:26 PM
"OpenAI lays groundwork for juggernaut IPO at up to US$1 trillion valuation" (tinyurl.com/2k6skuk3) JK--Aiming for 2027 IPO. The AI bubble might burst before then. Curious whether $1T valuation will be in the ballpark.
OpenAI lays groundwork for juggernaut IPO at up to US$1 trillion valuation
OpenAI is laying the groundwork for an initial public offering that could value the company at up to US$1 trillion, three people familiar with the matter said, in what could be one of the biggest IPOs...
tinyurl.com
November 4, 2025 at 7:54 PM
"Agents Rule of Two: A Practical Approach to AI Agent Security" (tinyurl.com/yvebs35x) JK--Good Meta discussion of specific constraints under which agentic AI apps can mitigate risk of prompt injection attacks.
Agents Rule of Two: A Practical Approach to AI Agent Security
We've developed the Agents Rule of Two. When this framework is followed, the severity of security risks is deterministically reduced.
tinyurl.com
November 4, 2025 at 5:06 PM
"Amazon Q3 Results: CEO Jassy Will Keep Spending To Meet AI Demand" (tinyurl.com/ye2bybu9) JK--Capacity expected to double by end of 2027.
Amazon Q3 Results: CEO Jassy Will Keep Spending To Meet AI Demand
Amazon.com plans to continue spending on capacity to meet demand for artificial intelligence and cloud products.
tinyurl.com
November 4, 2025 at 2:13 PM
"Big Tech Is Spending More Than Ever on AI and It’s Still Not Enough" (tinyurl.com/mn7xz34v) JK--You just know that the moment the rate of growth in business/consumer spending on AI-enabled apps/solutions begins to slow, there'll be a major freakout in the investment community.
Big Tech Is Spending More Than Ever on AI and It’s Still Not Enough
Meta, Alphabet, Microsoft and Amazon have all said they will increase spending in 2026. But investors have given mixed signals.
tinyurl.com
October 31, 2025 at 4:46 PM
"World Models in a Closed-Loop World" (world-in-world.github.io?utm_source=t...) JK--I like the research. I also like how they're visualizing this.
World-in-World: World Models in a Closed-Loop World
By grounding assessment in embodied task success instead of video metrics, World-In-World provides a principled yardstick for future research on generative world models in the context of embodiment.
world-in-world.github.io
October 23, 2025 at 3:37 PM
"Launching our new browser, ChatGPT Atlas" (tinyurl.com/39n8a2r2) JK--OK. But if your Mac laptop is running an Intel chip (like mine), you can't install it. Why not?
Launching our new browser, ChatGPT Atlas
Atlas is built with ChatGPT at its core so it can enhance your experience all across the web.
tinyurl.com
October 22, 2025 at 5:00 PM
"Is AI a bubble?" (tinyurl.com/3epcyrzv) JK--Not yet. "Capex to revenue ratios (6x) are high, exceeding both railroads (2x) and telecom (4x) at their peaks, but there are other indicators [not yet met]. Until two of these are met, the current risk, while large, is manageable."
🫧 Is AI a bubble?
A practical framework to test whether AI is building value - or inflating it.
tinyurl.com
October 17, 2025 at 6:42 PM
"Air Street Capital's eighth annual State of AI Report." Download it here: www.stateof.ai. 313 pages. Packed with good stuff.
State of AI Report 2025
The State of AI Report analyses the most interesting developments in AI. Read and download here.
www.stateof.ai
October 16, 2025 at 4:45 PM
"Sam Altman says ChatGPT is getting into erotica by the end of the year" (tinyurl.com/5n8u4ph2) JK--For so-called "verified adults" only. Juicier stuff if you can verify you're adulter than the rest. Exclusive top tier for those who can verify they're adulterers.
Sam Altman says ChatGPT is getting into erotica by the end of the year
Sam Altman announced that ChatGPT will roll out more extensive age-gating, including a new sexual feature.
tinyurl.com
October 15, 2025 at 2:48 PM
"I tested all of ChatGPT's new app integrations - here are the ones actually worth your time" (shorturl.at/WGvMF) JK--Promising. But you still need to vet the results closely to make sure it's satisfying your intention.
I tested all of ChatGPT's new app integrations - here are the ones actually worth your time
The future of ChatGPT with Spotify, Canva, and more shows promise, but there are still kinks to work out.
shorturl.at
October 14, 2025 at 6:54 PM
"State of LLMs in Late 2025" (shorturl.at/SA8ZD) JK--Good summary of technology, architecture, and practice at this time.
State of LLMs in Late 2025
Analyzing the current LLM landscape and its use-cases
shorturl.at
October 9, 2025 at 6:48 PM
"Welcome to AI's mega-blob era" (shorturl.at/uo6e3) JK--"The AI business is beginning to function like one giant dollar-eating, energy-sucking entity that makes chips, trains models and sketches utopias to justify its runaway costs."
Welcome to the mega-blob: AI firms are fusing into one big lump
OpenAI's AMD deal is the latest sign big AI firms are entangling supply chains, ownership and responsibility.
shorturl.at
October 9, 2025 at 6:32 PM
"Operationalizing Enterprise AI" (shorturl.at/JQIit) JK--"The very design of AI systems (trained on larger datasets, with growing parameters, reinforced by ecosystem effects) ensures that platforms will not stabilize for long."
Operationalizing Enterprise AI: Architecture Churn, Emergent Behaviors, and Strategic Foresight
A synthesis of insights from books, reports, and expert dialogues (that we have included in previous OODA Loop analysis or were in our research queue) all point to a central theme: enterprises should ...
shorturl.at
October 9, 2025 at 4:22 PM
"How long can Nvidia stay ahead of Chinese competition?" (shorturl.at/dp2xT) JK--Good chart showing how far ahead Nvidia's chips are, performance-wise, vis-a-vis US and overseas rivals.
October 9, 2025 at 4:13 PM
"OpenAI, Nvidia Fuel $1 Trillion AI Market With Web of Circular Deals" (shorturl.at/fuL89) JK--I'm starting to get that sense too.
OpenAI, Nvidia Fuel $1 Trillion AI Market With Web of Circular Deals
A wave of deals and partnerships are escalating concerns that the trillion-dollar AI boom is being propped up by interconnected business transactions.
shorturl.at
October 9, 2025 at 12:31 PM
"How Leaders Navigate Generative AI" (shorturl.at/8vpBS) JK--Very good diagrams on who's investing and otherwise partnering with who in this space.
How Leaders Navigate Generative AI
It's 'build baby build' against the world, and a rising chance of a recession. With a deep dive on AI leadership strategy.
shorturl.at
October 9, 2025 at 12:24 PM
"To Understand AI, Watch How It Evolves" (shorturl.at/lDBqi) Jk--Yeah, but how to explain stochastic gradient descent in such a way as not to put your audience sleep?
To Understand AI, Watch How It Evolves | Quanta Magazine
Naomi Saphra thinks that most research into language models focuses too much on the finished product. She’s mining the history of their training for insights into why these systems work the way they d...
shorturl.at
October 8, 2025 at 7:21 PM