Topics

Browse the writings on The Invariant by subject, idea, and recurring theme.

privacy 1

hard work 1

ideas 1

beginnings 1

artificial intelligence 2

machine learning 6

training data 1

data bias 2

algorithmic bias 2

AI safety 2

Elon Musk 1

Garden of Eden 1

Golem 1

collective unconscious 1

behavioural targeting 1

feedback loops 3

Docker 3

Python 5

VirtualBox 1

Boot2Docker 1

containers 1

reproducibility 2

dependency management 1

development environments 1

Bitcoin 2

Ethereum 1

proof of work 1

trustless systems 1

Zerocash 1

zk-SNARKs 1

cryptography 1

decentralisation 1

governance 2

data provenance 3

Git 2

CAP theorem 1

Eric Brewer 1

consistency 1

availability 1

partition tolerance 1

PACELC 1

clinical data 4

system design 1

bias-variance trade-off 1

support vector machines 1

decision trees 1

tree ensembles 1

clinical trials 1

patient recruitment 1

model evaluation 2

sensitivity 1

specificity 1

precision 1

recall 1

human review 2

SQL 1

MedCodes 1

AlexNet 1

transfer learning 1

Caffe 1

scikit-learn 2

stroke 1

diffusion MRI 1

ensemble learning 2

medical imaging 1

OpenAI 2

DeepMind 1

AlphaGo 2

Brexit 1

Babel 1

Genesis 11 1

artificial general intelligence 1

AI governance 3

compute 2

Docker Compose 1

Django 5

cloud computing 3

hot desking 1

software whitelisting 1

portable Python 1

reproducible environments 1

runtime 1

institutional IT 1

local training 1

distributed evaluation 1

gold standard 1

model persistence 1

Kubernetes 2

container orchestration 1

desired state 1

Pods 1

Services 1

Deployments 1

scheduling 1

reconciliation 1

distributed systems 2

research software 1

Prometheus 2

Grafana 2

monitoring 1

metrics 2

PromQL 1

alerting 1

information asymmetry 1

organisational silos 1

business intelligence 1

pattern recognition 1

decision support 3

internal tooling 1

natural language processing 3

recurrent neural networks 1

user interface automation 1

clinical informatics 1

career transition 1

linear regression 1

gradient descent 1

model training 1

prediction 1

optimisation 2

data 2

symbolic artificial intelligence 1

deep learning 2

neural networks 2

computer vision 2

reinforcement learning 2

chronic pain 1

back pain 1

pain perception 1

neuroplasticity 1

mind body 1

fear avoidance 1

John Sarno 1

competitive analysis 1

surrogate models 1

pricing 1

insurance 2

regression 1

market intelligence 1

Go 1

multithreading 1

multiprocessing 1

hyper-threading 1

GIL 1

goroutines 2

channels 2

backpressure 2

memory 2

selection bias 1

representative data 1

algorithmic fairness 2

scientific method 1

data governance 1

bias variance tradeoff 1

dataset shift 1

explainability 1

deployment context 1

model governance 5

explainable artificial intelligence 1

algorithmic transparency 1

LIME 1

model interpretation 1

human machine interaction 1

accountability 2

big data 1

missing data 1

counterfactuals 1

human judgement 3

data collection 1

attention economy 1

business incentives 1

emotional contagion 1

research ethics 1

personalisation 2

algorithmic governance 1

engagement 1

automation 6

tacit knowledge 1

lifelong learning 1

career resilience 1

human machine collaboration 1

interdisciplinary skills 1

network effects 1

platform economy 1

digital ecosystems 1

competitive strategy 1

marketplaces 1

TensorFlow 1

Keras 1

Jupyter 1

Anaconda 1

data pipelines 2

MLflow 1

PySpark 1

cross-validation 1

experiment tracking 1

model provenance 1

AI-native companies 1

business models 2

modular organisations 1

orchestration 1

digital platforms 1

model selection 1

deployment 1

predictive analytics 1

decision systems 2

MLOps 1

business processes 1

AI readiness 1

chatbots 1

pilot projects 1

organisational change 2

customer service 1

vendor strategy 1

AI strategy 1

operating models 1

executive accountability 1

data strategy 1

responsible AI 1

transparency 1

organisational design 1

process redesign 1

digital infrastructure 1

Estonia 1

digital literacy 1

GPT-2 1

Richard Sutton 1

the bitter lesson 1

AlphaGo Zero 1

language models 3

calculators 1

checkpoints 1

GPT-3 1

Common Crawl 1

data ownership 1

memorisation 1

digital labour 1

workflow automation 1

source of truth 1

PostgreSQL 3

audit trails 1

operational systems 1

process design 1

charity technology 1

Flutter 1

FastAPI 1

API design 1

business logic 1

shared state 1

client-server architecture 1

Celery 1

Redis 1

asynchronous processing 1

background jobs 1

idempotency 1

retry logic 1

workflow state 1

operational reliability 1

impostor phenomenon 1

Dunning-Kruger effect 1

software engineering 1

professional identity 1

self-assessment 1

technical education 1

burnout 1

XGBoost 2

EHCP 1

student monitoring 1

human in the loop 1

anomaly detection 1

educational data 1

false positives 1

longitudinal data 1

safeguarding 6

BART 1

Pegasus 1

summarisation 1

hierarchical summarisation 1

context windows 1

clinical abstraction 1

evidence preservation 1

case reports 1

document retrieval 1

export controls 1

semiconductors 1

China 2

Hephaestus 1

Tony Stark 1

scarcity 1

supercomputing 1

retrieval augmented generation 9

RAG 2

large language models 7

policy retrieval 1

grounding 4

external knowledge 1

document governance 1

evidence 2

local policy 1

ChatGPT 2

Aizen 1

Bleach 2

Nietzsche 1

Blade Runner 1

Tower of Babel 1

AI agents 4

projection 1

fast.ai 1

Cerbos 3

LlamaIndex 2

role-based access control 1

least privilege 1

policy enforcement 1

access control 1

human on the loop 1

OpenSearch 1

document chunking 1

hierarchical retrieval 1

document parsing 1

deterministic software 1

retrieval-augmented generation 1

software architecture 1

QLoRA 2

LoRA 3

fine-tuning 1

prompt engineering 2

model adaptation 1

GPU memory 2

human oversight 3

LLM serving 2

continuous batching 1

dynamic batching 1

fixed batching 1

token streaming 1

autoregressive generation 1

GPU utilisation 1

inference scheduling 1

Yhwach 1

AI systems 1

tool use 3

multimodality 1

embodiment 1

machine awareness 1

sovereignty 1

Ghost in the Shell 1

Dune 1

The Matrix 1

Tiamat 1

Revelation 1

quantisation 2

distillation 1

mixture of experts 4

AI infrastructure 1

Langfuse 2

OpenLineage 1

Pydantic 1

model lineage 1

traceability 1

GPT-4 1

Code Interpreter 1

Assistants API 1

LangGraph 1

AI interfaces 1

PagedAttention 1

KV cache 1

vLLM 1

inference optimisation 1

memory management 2

data quality 3

systems thinking 1

institutional knowledge 1

operational knowledge 1

observability 2

technical debt 5

epistemic debt 2

software reliability 1

database releases 1

quality control 1

HPLC 1

validation 1

release gates 1

hotfixes 2

operational debt 2

fitness for purpose 1

data conformance 1

synthetic data 2

quality dashboards 1

data lineage 3

source contracts 1

data transformation 1

Conway's Law 1

organisational knowledge 1

DeepSeek-R1 1

open weights 2

refusal 1

alignment 1

Mark 5 1

local models 1

distribution 1

data extraction 1

batch processing 1

checkpointing 1

manifests 1

checksums 1

system resilience 1

concurrency 1

parallelism 1

processes 1

threads 1

Rust 1

evaluation 1

model traces 1

agent evaluation 1

trajectory evaluation 1

sensitive data 1

restricted environments 1

evaluation harnesses 1

Model Context Protocol 1

MCP 1

APIs 1

interoperability 1

authorisation 1

prompt injection 1

tool poisoning 1

South Korea 2

DeepSeek 2

robotics 1

KOSPI 1

fertility 2

AI investment 1

OpenTelemetry 1

distributed tracing 1

readiness 1

operational health 1

AI observability 1

transformers 1

attention mechanisms 1

embeddings 1

RoPE 1

model architecture 1

LLM infrastructure 1

release governance 1

decision authority 1

risk acceptance 1

exceptions 1

provenance 1

Children of Men 1

gender 1

housing 1

inflation 1

Keynesian 1

Austrian 1

labour 1

domain 1

website 1

theinvariant.uk 1

Kimi K3 1

Moonshot AI 1

GPU infrastructure 1

inference economics 1