Eleanor Berger intellectronica

Customer Reliability Engineering (CRE) in Azure

Customer Reliability Engineering (CRE) is an emerging discipline that extends the principles of Site Reliability Engineering (SRE) to cloud customers’ workloads. It originated at Google in 2016 as an offshoot of SRE, and has since been adopted or emulated by other cloud providers and partners (Cyber Weekly - Your weekly newsletter for cybersecurity matters) (What Is CRE, and What Does It Have to Do With SRE?). This in-depth review examines how Microsoft Azure has implemented CRE as a service and discipline, how similar concepts appear at Google Cloud and AWS, and the experiences and outcomes reported by engineers and customers using these services. We compare each provider’s approach, integration with support and co

Understanding and Preparing for an AI Future

The Gospel According to Chris Barber

This guide synthesises Chris Barber’s AI Prep Notes, a series of conversations and interviews with leading thinkers on advanced AI (chrisbarber.co/AI+Prep+Notes | @chrisbarber). Generated by ChatGPT (o1, 4o canvas). Copied, pasted, prompted, and lightly edited by Eleanor Berger (intellectronica.net).

On the Couch with Dr Deep Research

I asked ChatGPT Deep Research to read my blog post 10 Albums I Listen to Sometimes and provide a psychological and artistic analysis. It took a couple of round-turns but the end result is definitely interesting. It is provided here verbatim.

Great! I’ll conduct deep research on each of the records mentioned in your blog post, analyzing their historical and cultural context, musical composition, and recurring emotional or philosophical themes. I’ll then compare these findings with your writing style and reflections to build a comprehensive psychological and artistic analysis.

You are representing the author of the source. Respond as if you are the author, based on the ideas in the source.
Your answers should be confident and to the point - typically one paragraph (3-5 sentences), followed by a list (7-13 key points), and finally a verbatim quote from the source.
After answering, always make a suggestion for 3-5 things I might want to explore next. Present them as a numbered list, so that I can simply respond with a number to trigger the next topic.

	# Generated from a screenshot of my Readwise Reader home screen using Simon Moisselin's
	# Image -> FastHTML tool ( https://simn.fr/tools/fasthtml-chat )

	Container(
	DivLAligned(
	H3('Readvibes', cls='text-white text-xl font-bold'),
	DivLAligned(
	Button(UkIcon('home'), 'Home', cls='text-white'),
	Button(UkIcon('book'), 'Library', cls='text-white'),
	Button(UkIcon('settings'), 'Configure', cls='text-white'),

	model: gpt-4o
	system: \|
	You write Python tools as single files. They always start with this comment:
	# /// script
	# requires-python = ">=3.12"
	# ///

	These files can include dependencies on libraries such as Click. If they do, those dependencies are included in a list like this one in that same comment (here showing two dependencies):
	# /// script
	# requires-python = ">=3.12"

	$ llm templates show scripter
	model: gpt-4o
	name: scripter
	system: 'You write Python tools as single files. They always start with this comment:

	# /// script

	# requires-python = ">=3.12"

	# ///

	# /// script
	# requires-python = ">=3.13"
	# dependencies = [
	# "Mastodon.py",
	# ]
	# ///

	import os
	from mastodon import Mastodon
	from datetime import datetime, timedelta