Felix Barbalet
@felixbarbalet.com
software engineering w/ #clojure, #data, #electronics, #photography, #infosec and #economics. My longer form writing is @ https://felixbarbalet.com
https://keyoxide.org/hkp/535FFD2607AFCB2C97831159CD1D679C615C17C8
https://keyoxide.org/hkp/535FFD2607AFCB2C97831159CD1D679C615C17C8
9/10. We Confuse the Process with the Facts About the Process
Your execution engine shouldn't just run the workflow; it must record every step as an immutable fact. We use Datomic not just for the domain data, but for the orchestration itself—workflows, stages, tasks. These are entities.
Your execution engine shouldn't just run the workflow; it must record every step as an immutable fact. We use Datomic not just for the domain data, but for the orchestration itself—workflows, stages, tasks. These are entities.
December 9, 2025 at 7:34 AM
9/10. We Confuse the Process with the Facts About the Process
Your execution engine shouldn't just run the workflow; it must record every step as an immutable fact. We use Datomic not just for the domain data, but for the orchestration itself—workflows, stages, tasks. These are entities.
Your execution engine shouldn't just run the workflow; it must record every step as an immutable fact. We use Datomic not just for the domain data, but for the orchestration itself—workflows, stages, tasks. These are entities.
8/10. The Knowledge Graph as Constraint
Many see the Knowledge Graph solely as a source for Retrieval-Augmented Generation (RAG). This misses half the value. In a high-stakes system, the KG is also a constraint. It provides the deterministic boundaries within which the LLM is allowed to operate.
Many see the Knowledge Graph solely as a source for Retrieval-Augmented Generation (RAG). This misses half the value. In a high-stakes system, the KG is also a constraint. It provides the deterministic boundaries within which the LLM is allowed to operate.
December 8, 2025 at 6:01 AM
8/10. The Knowledge Graph as Constraint
Many see the Knowledge Graph solely as a source for Retrieval-Augmented Generation (RAG). This misses half the value. In a high-stakes system, the KG is also a constraint. It provides the deterministic boundaries within which the LLM is allowed to operate.
Many see the Knowledge Graph solely as a source for Retrieval-Augmented Generation (RAG). This misses half the value. In a high-stakes system, the KG is also a constraint. It provides the deterministic boundaries within which the LLM is allowed to operate.
7/10. The Tyranny of Structure vs. The Power of Data
We often force information into rigid schemas too early. A well-designed engine stores intermediate results—the outputs of LLM calls—as simple, serialized data (like EDN payloads).
We often force information into rigid schemas too early. A well-designed engine stores intermediate results—the outputs of LLM calls—as simple, serialized data (like EDN payloads).
December 6, 2025 at 10:19 PM
7/10. The Tyranny of Structure vs. The Power of Data
We often force information into rigid schemas too early. A well-designed engine stores intermediate results—the outputs of LLM calls—as simple, serialized data (like EDN payloads).
We often force information into rigid schemas too early. A well-designed engine stores intermediate results—the outputs of LLM calls—as simple, serialized data (like EDN payloads).
6/10. Isolating the Plumbing
We often complect the core logic of a task (the prompt engineering, the tool definitions) with the accidental complexity of its execution (error handling, retries, timing, validation). This entanglement guarantees rigidity.
We often complect the core logic of a task (the prompt engineering, the tool definitions) with the accidental complexity of its execution (error handling, retries, timing, validation). This entanglement guarantees rigidity.
December 5, 2025 at 9:57 PM
6/10. Isolating the Plumbing
We often complect the core logic of a task (the prompt engineering, the tool definitions) with the accidental complexity of its execution (error handling, retries, timing, validation). This entanglement guarantees rigidity.
We often complect the core logic of a task (the prompt engineering, the tool definitions) with the accidental complexity of its execution (error handling, retries, timing, validation). This entanglement guarantees rigidity.
5/10. Dynamic Adaptation vs. Static Plans
A static workflow definition is a plan made when you know the least. Agentic systems, by definition, adapt. The path depends on the results of the previous stochastic step. The engine must support dynamic stages guided by "assessors."
A static workflow definition is a plan made when you know the least. Agentic systems, by definition, adapt. The path depends on the results of the previous stochastic step. The engine must support dynamic stages guided by "assessors."
December 3, 2025 at 9:25 AM
5/10. Dynamic Adaptation vs. Static Plans
A static workflow definition is a plan made when you know the least. Agentic systems, by definition, adapt. The path depends on the results of the previous stochastic step. The engine must support dynamic stages guided by "assessors."
A static workflow definition is a plan made when you know the least. Agentic systems, by definition, adapt. The path depends on the results of the previous stochastic step. The engine must support dynamic stages guided by "assessors."
4/10. The Necessity of an Immutable Ledger
In high-stakes environments, "It usually works" is malpractice. When an LLM makes a decision, you must know why. This requires more than logs. It requires an immutable, accretive Execution Ledger.
In high-stakes environments, "It usually works" is malpractice. When an LLM makes a decision, you must know why. This requires more than logs. It requires an immutable, accretive Execution Ledger.
December 2, 2025 at 7:34 AM
4/10. The Necessity of an Immutable Ledger
In high-stakes environments, "It usually works" is malpractice. When an LLM makes a decision, you must know why. This requires more than logs. It requires an immutable, accretive Execution Ledger.
In high-stakes environments, "It usually works" is malpractice. When an LLM makes a decision, you must know why. This requires more than logs. It requires an immutable, accretive Execution Ledger.
3/10: Composition over Command: The Functional Dataflow
If your architecture diagrams look like flowcharts, you're building a brittle system. Flowcharts are about control and sequence—concepts that stochastic components actively resist. Instead, design for dataflow.
If your architecture diagrams look like flowcharts, you're building a brittle system. Flowcharts are about control and sequence—concepts that stochastic components actively resist. Instead, design for dataflow.
December 1, 2025 at 6:01 AM
3/10: Composition over Command: The Functional Dataflow
If your architecture diagrams look like flowcharts, you're building a brittle system. Flowcharts are about control and sequence—concepts that stochastic components actively resist. Instead, design for dataflow.
If your architecture diagrams look like flowcharts, you're building a brittle system. Flowcharts are about control and sequence—concepts that stochastic components actively resist. Instead, design for dataflow.
2/10 Facts, Time, and the Peril of Caching Hallucinations
Stochastic systems demand memoization. But caching an LLM output is fraught. What if the underlying facts in your knowledge graph change? A naive cache returns a stale lie. Information is situated.
Stochastic systems demand memoization. But caching an LLM output is fraught. What if the underlying facts in your knowledge graph change? A naive cache returns a stale lie. Information is situated.
November 29, 2025 at 10:19 PM
2/10 Facts, Time, and the Peril of Caching Hallucinations
Stochastic systems demand memoization. But caching an LLM output is fraught. What if the underlying facts in your knowledge graph change? A naive cache returns a stale lie. Information is situated.
Stochastic systems demand memoization. But caching an LLM output is fraught. What if the underlying facts in your knowledge graph change? A naive cache returns a stale lie. Information is situated.
1/10: The Illusion of Control
We crave determinism. We build orchestration engines that execute Directed Acyclic Graphs with atomic precision. Then, we insert a stochastic black box - an LLM - into a stage and expect the precision to hold. This is a fundamental misunderstanding.
We crave determinism. We build orchestration engines that execute Directed Acyclic Graphs with atomic precision. Then, we insert a stochastic black box - an LLM - into a stage and expect the precision to hold. This is a fundamental misunderstanding.
November 28, 2025 at 9:57 PM
1/10: The Illusion of Control
We crave determinism. We build orchestration engines that execute Directed Acyclic Graphs with atomic precision. Then, we insert a stochastic black box - an LLM - into a stage and expect the precision to hold. This is a fundamental misunderstanding.
We crave determinism. We build orchestration engines that execute Directed Acyclic Graphs with atomic precision. Then, we insert a stochastic black box - an LLM - into a stage and expect the precision to hold. This is a fundamental misunderstanding.
Today I realised another one of the things I love about #Clojure is the number of powerful patterns embodied not only in the core, but in the libraries and the ecosystem. Learning by reading others code means I need "good company", and Clojure (and it's ecosystem) delivers.
November 24, 2025 at 4:13 AM
Today I realised another one of the things I love about #Clojure is the number of powerful patterns embodied not only in the core, but in the libraries and the ecosystem. Learning by reading others code means I need "good company", and Clojure (and it's ecosystem) delivers.
Happy Friday !
Chapter 1 - The Anomaly at Farside
The silence of Farside Station was not a natural occurrence, but a meticulously engineered state. It was an absolute, profound quiet that existed nowhere on the turbulent sphere of Earth.
gist.github.com/xlfe/0b7dc10... #scifi #novel
Chapter 1 - The Anomaly at Farside
The silence of Farside Station was not a natural occurrence, but a meticulously engineered state. It was an absolute, profound quiet that existed nowhere on the turbulent sphere of Earth.
gist.github.com/xlfe/0b7dc10... #scifi #novel
A new, previously undiscovered Sci Fi Novel
A new, previously undiscovered Sci Fi Novel. GitHub Gist: instantly share code, notes, and snippets.
gist.github.com
October 17, 2025 at 6:11 AM
Happy Friday !
Chapter 1 - The Anomaly at Farside
The silence of Farside Station was not a natural occurrence, but a meticulously engineered state. It was an absolute, profound quiet that existed nowhere on the turbulent sphere of Earth.
gist.github.com/xlfe/0b7dc10... #scifi #novel
Chapter 1 - The Anomaly at Farside
The silence of Farside Station was not a natural occurrence, but a meticulously engineered state. It was an absolute, profound quiet that existed nowhere on the turbulent sphere of Earth.
gist.github.com/xlfe/0b7dc10... #scifi #novel
Reposted by Felix Barbalet
Reposted by Felix Barbalet
Choosing a microcontroller: lcamtuf.substack.com/p/choosing-a...
Choosing a microcontroller
RP2040, ESP32, AVR, PIC, STM32...? When it comes to hobby projects, there's plenty of choice and just as much zealotry.
lcamtuf.substack.com
February 8, 2025 at 5:40 PM
Choosing a microcontroller: lcamtuf.substack.com/p/choosing-a...
What's truly amazing about #DeepSeek R1? It's that it's been released as completely #OpenSource. But why would DeepSeek choose to Open Source the product of it's innovation? www.linkedin.com/pulse/deepse... #economics #strategy
DeepSeek Just Shook the AI World: Why Open-Sourcing R1 Could Be the Smartest Move in the AI Arms Race 🚀
What's truly amazing about DeepSeek's R1? It's not that it's on par with the (up to now) leading models from the US - it's that it's been released as completely Open Source. When we're in the midst of...
www.linkedin.com
January 28, 2025 at 10:22 PM
What's truly amazing about #DeepSeek R1? It's that it's been released as completely #OpenSource. But why would DeepSeek choose to Open Source the product of it's innovation? www.linkedin.com/pulse/deepse... #economics #strategy
Just having a look around the uLisp site and was impressed to see the detailed performance benchmarks including power www.ulisp.com/show?36M3 - I added time * average power as some boards are really quick, but high power o_O
#uLisp #embedded #benchmarks #microcontroller #lisp #programming #lowpower
#uLisp #embedded #benchmarks #microcontroller #lisp #programming #lowpower
January 22, 2025 at 12:53 AM
Just having a look around the uLisp site and was impressed to see the detailed performance benchmarks including power www.ulisp.com/show?36M3 - I added time * average power as some boards are really quick, but high power o_O
#uLisp #embedded #benchmarks #microcontroller #lisp #programming #lowpower
#uLisp #embedded #benchmarks #microcontroller #lisp #programming #lowpower
Bios update with flashrom went perfectly using a UM232H with a level shifter and LM317 to get the 1.8v required for the SPI BIOS chip
#bios #coreboot #electronics
#bios #coreboot #electronics
January 13, 2025 at 11:18 PM
Bios update with flashrom went perfectly using a UM232H with a level shifter and LM317 to get the 1.8v required for the SPI BIOS chip
#bios #coreboot #electronics
#bios #coreboot #electronics
A cost/benefit framework provides a valuable lens for prioritizing information management efforts and evaluating the potential of information assets. #InfoEconomics #CostBenefitAnalysis #DataStrategy
January 6, 2025 at 2:31 PM
A cost/benefit framework provides a valuable lens for prioritizing information management efforts and evaluating the potential of information assets. #InfoEconomics #CostBenefitAnalysis #DataStrategy
IT is not just a support function for information management; it's an integral component. Aligning IT strategy with information strategy is critical for success. #ITStrategy #InfoManagement #BusinessAlignment
January 3, 2025 at 2:31 PM
IT is not just a support function for information management; it's an integral component. Aligning IT strategy with information strategy is critical for success. #ITStrategy #InfoManagement #BusinessAlignment
While security is paramount, overly restrictive information access can hinder productivity and innovation. Finding the right balance is a key challenge for organizations. #InformationAccess #KnowledgeSharing #Productivity
January 2, 2025 at 2:31 PM
While security is paramount, overly restrictive information access can hinder productivity and innovation. Finding the right balance is a key challenge for organizations. #InformationAccess #KnowledgeSharing #Productivity
Effective knowledge sharing within an organization is crucial for leveraging collective expertise and reducing key-person risk. IT systems play a vital role in enabling this. #KnowledgeManagement #Collaboration #InfoSharing
January 1, 2025 at 2:31 PM
Effective knowledge sharing within an organization is crucial for leveraging collective expertise and reducing key-person risk. IT systems play a vital role in enabling this. #KnowledgeManagement #Collaboration #InfoSharing
Maximizing data value involves enhancing data quality, accessibility, and ease of use. These factors directly impact the return on investment in data initiatives. #DataQuality #DataAccessibility #DataUsability
December 31, 2024 at 2:31 PM
Maximizing data value involves enhancing data quality, accessibility, and ease of use. These factors directly impact the return on investment in data initiatives. #DataQuality #DataAccessibility #DataUsability
Reactive "blunder funding" (investing after a mistake) is a suboptimal approach to information management. Proactive, strategic investment yields better long-term results. #InformationInvestment #StrategicPlanning #DataBudget
December 30, 2024 at 2:31 PM
Reactive "blunder funding" (investing after a mistake) is a suboptimal approach to information management. Proactive, strategic investment yields better long-term results. #InformationInvestment #StrategicPlanning #DataBudget
Information exists in two primary forms: data (stored in systems) and knowledge (embodied in people). Each requires distinct management strategies. #DataVsKnowledge #InfoManagement #ITStrategy
December 27, 2024 at 2:31 PM
Information exists in two primary forms: data (stored in systems) and knowledge (embodied in people). Each requires distinct management strategies. #DataVsKnowledge #InfoManagement #ITStrategy
Information theory highlights the connection between reduced uncertainty and value creation. Effectively using information leads to better decisions and outcomes. #InformationTheory #DecisionMaking #DataDriven
December 26, 2024 at 2:31 PM
Information theory highlights the connection between reduced uncertainty and value creation. Effectively using information leads to better decisions and outcomes. #InformationTheory #DecisionMaking #DataDriven
Information assets present both potential and realized value, alongside inherent risks. A balanced approach considers both value extraction and risk mitigation. #DataRisk #InfoEconomics #StrategicAsset
December 25, 2024 at 2:31 PM
Information assets present both potential and realized value, alongside inherent risks. A balanced approach considers both value extraction and risk mitigation. #DataRisk #InfoEconomics #StrategicAsset