
Founder & CEO @norwoodsystems – next-gen agentic enablement systems. Opinions are my own.
Perth, Western Australia
Joined April 2008
- Tweets4.9K
- Following2.9K
- Followers1.2K
- Likes28.7K
Paul Ostergaard retweeted
AI is by far the number one technology in human history, and this will soon become apparent once it reaches superintelligence.
Paul Ostergaard retweeted
Replying to @itsdurreadan
Incredibly poor form on Apple’s part - as you say all the input data is already on the earlier watch models. Given most of these metrics are readable via Apple Health ‘someone’ just going to have launch an app that embarrasses Apple supporting these earlier hardware models.
Nokia used to block access to their phones’ audio subsystems (to my great chagrin 20 years ago when my team and I were building the first mobile VoIP apps ever) for reasons unrelated to customer utility. I sincerely hope that Apple is not turning into Nokia, the total antithesis of who it used to be. @johnternus @tim_cook
Paul Ostergaard retweeted
I would print and put this opinion from @sama on the wall to read it every day! This is exactly why we must resist and fight back against those who want to be the AI overlords! (There is at least an aspiring one already!)
"A lot of the talk about safety concerns is well-founded, and then a lot of it is about people that just really, even if it's slightly subconscious, wanna concentrate power.
I don't think anyone should want to live in a world of AI overlords, where someone is making decisions for all of the future, and in exchange for a cure for cancer, which obviously is a wonderful thing, we collectively cede all agency.
I am terrified of a world where the very real fears of AI are used as a way to say, “Only this small group of people can have it because it's too dangerous, and only they understand it. But don't worry, they're gonna make the right decisions for all of us.”
Sam Altman: “Concentration of power with AI is a terrifying thing.”
"A lot of the talk about safety concerns is well-founded, and then a lot of it is about people that just really, even if it's slightly subconscious, wanna concentrate power.
I don't think anyone should want to live in a world of AI overlords, where someone is making decisions for all of the future, and in exchange for a cure for cancer, which obviously is a wonderful thing, we collectively cede all agency.
I am terrified of a world where the very real fears of AI are used as a way to say, “Only this small group of people can have it because it's too dangerous, and only they understand it. But don't worry, they're gonna make the right decisions for all of us.”
I was a child of the internet. There were no rules. It was amazing, and I think it's critical we preserve that spirit with AI."
Paul Ostergaard retweeted
GPT-6 news: @OpenAI engineer Roon indirectly confirmed that GPT-6 is coming soon!
He also hinted that GPT-6 will be a *substantial* upgrade over the latest model GPT-5.6.
My prediction:
GPT-6 is coming within the next 8 weeks. It will be a large model in the 6T to 10T class.
Paul Ostergaard retweeted
you must see LLMs as intelligence amplifiers and their amplification depends on your actual intelligence
so, if your IQ is 135+, with AI you can perform at the current AI maximum and it feels like real magic
if you’re 120-135, you get a good 50% buff
If you’re 105-120, you get a 25%
90-105, nothing changes
below 90 it makes you more stupid
Paul Ostergaard retweeted
if you're not operating at near-max intensity, you're already conceding to someone who is
progress isn't really a mood -- it’s a physics problem, and momentum decides who owns the future.
Paul Ostergaard retweeted
#1 trait for founders: resilience. So much is timing, and it comes down to your ability to keep trying just one more time, then one more time, until it finally hits
Paul Ostergaard retweeted
Congratulations - perhaps the last time a human prevails at programming!
Looking forward to diving deep at specific problems :)
Paul Ostergaard retweeted
I let this pass but shouldn't have.
Been tracking adoption of startups for a very long time.
ICQ was released November 1, 1996, for instance, and by the time I joined it in mid December it had 65,000 and was the fastest growing service up to then.
A million new users an hour is insane growth.
And people are shorting NVIDIA.
Makes no sense at all.
Paul Ostergaard retweeted
It seems @OpenAI is very serious about building its own robots! Of course this vertical integration makes perfect strategic sense-controlling both the intelligence and embodiment-robots with AGI will likely constitute at least quarter of the world’s economy within a decade.
Our team at @OpenAI is hiring a **Mechanical Architect, Robotics**. We are looking for senior candidates with experience designing complex robotic systems.
Please share with folks you think would be great!
Paul Ostergaard retweeted
Deep stuff! We uncovered a startling link between #entropy—a bedrock concept in #physics—and how #AI can keep discovering new ideas without stagnating. In an era where reasoning models can reflect on problems for days at a time (rather than generating quick, single-step solutions), our study shows how semantic entropy (the spread of meanings) and structural entropy (how evenly its links between concepts generated by the AI are distributed) together hold the secret to ongoing exploration. Specifically, we measured structural entropy using Von Neumann graph entropy (applied to the adjacency Laplacian), while semantic entropy came from a similarity-based embedding deep language embedding matrix. The key insight? Although semantic entropy consistently outpaces structural entropy, they remain in a near-critical balance—fueling "surprising edges" that introduce relationships between distant concepts. This mirrors physical systems on the brink of a phase transition, where a little bit of "disorder" keeps the process dynamic yet avoids chaos. The result is an AI that doesn’t just keep pace with known solutions but actively creates new pathways of thought over extended “thinking” sessions.
As reasoning models become ever more capable—undertaking extended, multi-day "thought processes"—understanding fundamental principles is crucial. By weaving these insights into reinforcement learning strategies, we can reward models not just for correctness, but for venturing into novel conceptual ground. This opens the door to AI systems that actively cultivate new insights, rather than settling into narrow patterns or endlessly rehashing the same knowledge.
Going Deeper
When physicists describe entropy, they refer to the measure of "disorder" in a system: the number of ways particles can rearrange without altering the system’s energy. Yet entropy transcends molecules and heat. In this research, it emerges as the engine that drives AI reasoning models to keep generating fresh ideas over extended periods. By analyzing agentic graph reasoning, we computed Structural Entropy (using Von Neumann graph entropy of the normalized Laplacian) to measure how evenly a knowledge graph’s connections spread out. In parallel, we assessed Semantic Entropy, built from a semantic adjacency matrix of cosine similarities between node embeddings. Remarkably, semantic entropy persistently outran its structural counterpart, indicating that the AI remains open to far-reaching conceptual links while sustaining a stable fraction of "surprising edges."
This dynamic reflects self-organized criticality—a state where systems hover between rigid order and random chaos. Much like a sand pile teetering on the edge of collapse, the AI preserves enough organizational structure to remain coherent, yet stays flexible enough to generate unexpected leaps in meaning. The fraction of "surprising edges" remains stable, offering evidence that the model naturally integrates new, distant ideas without toppling into confusion. Ultimately, this deep linkage between physics and AI may become a cornerstone for how we build machines that never stop learning, tapping into the same principle that governs how atoms move in a gas—entropy as the driving force behind endless possibility.
Paul Ostergaard retweeted
This isn't quite true. Test-time compute helps when verification is easier than generation (e.g., sudoku), but if the task is "When was George Washington born?" and you don't know, no amount of thinking will get you to the correct answer. You're bottlenecked by verification.
A crucial point that everyone should be internalizing: in the age of test-time search, it's pretty much always possible to reach any level of capability by simply expending more compute.
So it’s not just about, "can you do it?". The key is how efficiently you can do it. Intelligence is not just the ability to demonstrate high skill. It's the efficiency with which you acquire and deploy these skills. Both data efficiency and compute efficiency.
Going forward we will report ARC-AGI scores on a 2D plot – score on one axis and cost per task on the other. In general, this is the only sane way to report performance for test-time reasoning models.
Paul Ostergaard retweeted
Watch our latest podcast with Norwood Systems (NOR) CEO Paul Ostergarrd CEO aul Ostergaard youtu.be/Xz0-u7z-0JE?si=XQcd… via @YouTube
@norwoodsystems @paulostergaard
#asx #asxNews #asx200 #ausbiz #tradingday
Paul Ostergaard retweeted
Another huge week in AI Agents 🧵
I summarized everything announced by AgentOps, NVIDIA, AutoGen, Langchain, Qodo, Box, Graphite, Mastra AI, Stripe, and more.
Here's everything you need to know and how to make sense of it:
(save for later)
Paul Ostergaard retweeted
It's getting harder to tell which replies here on X are from humans and which ones are from AI's. One tip? The AI's answer way too fast. :-)
What's your policy? Do you block AI answer accounts? Or do you hide them? Or do you just allow them as long as they are on point?
Paul Ostergaard retweeted
Wow, this is the „feel the AGI“- Moment for robotics for me.
Boston dynamics humanoid robots movement looks so clean it’s incredible.
The robot can walk, run, crawl and so much more. I never have thought to see something like this already this year.
I honestly can’t imagine what 2026 has for us!
Paul Ostergaard retweeted
This is the larval form of a $3 trillion company. It's very different looking, but not unrecognizable. In fact Y Combinator specializes in recognizing and helping it. (image via @JonErlichman)





























