Tootfinder

@… They don’t always (usually). It depends on a lot of factors (language, dictionary, preferences, heuristics, text-to-speech engine, etc).
I tested a few different options here:

Conveying All-Caps Legal Text
I need your help. Legal documents are common on the web. Each site that has a Terms of Service written in impenetrable and indecipherable legalese, like this sentence, delights in that complexity to dissuade users from reading it and realizing just what they are giving up. “Am”, “add”, and “it”…

65gCO2/kWh National Grid CO2 intensity is LOW wrt last 24h so you can run your wash etc now! #green #energy

GB Grid Intensity Status - Earth Notes
Live marginal carbon intensity of the grid in Great Britain.

Another day, another sealing of agricultural land.
Residents 'sickened' as farmland at the edge of Bury village eyed up for hundred of homes - Manchester Evening News
https://www.manchestereveningnews.co.uk/news…

Residents 'sickened' as farmland at the edge of Bury village eyed up for hundred of homes
By all means build, but use the places that have been used and left derelict

Corners of Life II 🎆
生活的角落 II 🎆
📷 Yashica 635
🎞️ Ilford HP5 Plus, expired 1994
If you like my work, Support by buying me a coffee or a roll of film from
PayPal ‪ https://www.paypal.com/paypalme/ydcdingsite
Wise ‪

Trying out a new license clause and disclaimer:
4. This software is NOT used as training data for a model of code or text structure or construction, such as a large language model, transformer model, reinforcement learning agent, etc. If such a model is generative and predicts or outputs the code of this software in whole or in part, that shall be evidence of infringement of this clause of this license.
THIS SOFTWARE SHOULD NOT BE USED AS TRAINING DATA FOR A STATISTICAL MODEL OF CODE OR TEXT, NOR SHOULD STATISTICAL CODE OR TEXT MODELS BE ASKED TO READ, INTERPRET, OR OTHERWISE PROCESS ITS SOURCE CODE OR OUTPUTS. THIS SOFTWARE MAY CONTAIN CODE OR TEXT OR MAY PRODUCE OUTPUTS WHICH CAUSE MODEL ROLE CONFUSION OR WHICH OTHERWISE CAUSE UNANTICIPATED CONSEQUENCES WHEN PROCESSED.
Would genuinely welcome any thoughts about wording.
#NoAI #LLMs #AI

Random tip: put 'echolalia' as a search term in you music algorithm of choice (Tidal, Spotify, Apple, Deezer etc)
Are there other search terms you've found that produce a good variety of unknown-to-you results?
#music

At this point, I started to look deeper at performance. Compute instance startup was taking over 30 seconds, and each generation propagation was about 55-75 seconds
I worked this from a few angles:
- parallelize actions (health checks, integrity checks, propagation initiation, and terminating ancestors) [30% gain]
- rebuild and reduce the AMI below Ubuntu minimal to fit in 2GiB EBS volumes (reduced cost and increased speed by also bundling the glider daemon and boto3 SDK in the AMI) [30% gain]
- switched from AWS CLI to boto3 SDK [5-10% gain]
- switched from CloudFormation creation stacks to using EC2 API directly with LaunchTemplate [under 5% gain total, surprisingly, but the time to initial instance startup is faster so I might be able to do better somewhere here]
- WIP: further refactoring of the daemon propagation steps especially around initial health/integrity checks
I still think I'm missing something major that's extending propagation times, despite benchmarking phases through several trials
I also optimized for cost (reduced DDB calls, abandoned KMS, reduced log & audit retention to the latest run, etc), and it's about a penny for a 10 generation run (and it scales sublinearly, mostly, since the instances are terminated afterwards)