mattcollins

6 ms·

1.
United Arab Emirates to quit oil cartel OPEC(bbc.co.uk)
15 points by mattcollins 5mo ago|1 comments
2.
A Timeline to China Blocking Meta's $2B Manus Acquisition (Built Using Manus)(metamanus-rsbcnkpx.manus.space)
3 points by mattcollins 5mo ago|0 comments
3.
by mattcollins 7mo ago
On the other hand, AI coding tools make it relatively easy to set and apply policies that can help with this sort of thing. I like to have something like the following in AGENTS.md: ## Guiding Principles - Optimise for long-term maintainabi
4.
US Justice Department releasing more than three million pages from Epstein files(bbc.co.uk)
41 points by mattcollins 8mo ago|13 comments
5.
by mattcollins 11mo ago
Results from some further tests here: https://www.improvingagents.com/blog/toon-benchmarks
6.
by mattcollins 11mo ago
FWIW, I ran a test comparing LLM accuracy with TOON versus JSON, CSV and a variety of other formats when using them to represent tabular data: https://www.improvingagents.com/blog/is-toon-good-for-table-... I've o
7.
by mattcollins 1y ago
This is a follow-up to previous work looking at which format of TABULAR data LLMs understand best: https://www.improvingagents.com/blog/best-input-data-format-... (There was some good discussion on Hacker News around t
8.
Which Nested Data Format Do LLMs Understand Best? JSON vs. YAML vs. XML vs. MD(improvingagents.com)
2 points by mattcollins 1y ago|1 comments
9.
by mattcollins 1y ago
Here you go: https://www.improvingagents.com/blog/best-input-data-format-...
10.
by mattcollins 1y ago
Author here. This has made me chuckle several times - thanks!
11.
by mattcollins 1y ago
I did a small test with just a couple of formats and something like 100 records, saw that the accuracy was higher than I wanted, then increased the number of records until the accuracy was down to 50%-ish (e.g. 100 -> 200 -> 500 ->
12.
by mattcollins 1y ago
I'm the person who ran the test. To hopefully clarify a bit... I intentionally chose input data large enough that the LLM would be scoring in the region of 50% accuracy in order to maximise the discriminative power of the test.
13.
by mattcollins 1y ago
I'm the person who ran the test. To explain the 60% a bit more... With small amounts of input data, the accuracy is near 100%. As you increase the size of the input data, the accuracy gradually decreases. For this test, I intentionally
14.
by mattcollins 1y ago
I'm the person who ran the test. The context I used in the test was pretty large. You'll see much better (near 100%) accuracy if you're using smaller amounts of context. [I chose the context size so that the LLM would be scor
15.
by mattcollins 1y ago
"This feature is available to all customers, meaning anyone can enable this today from the Cloudflare dashboard." https://blog.cloudflare.com/control-content-use-for-ai-train...
16.
by mattcollins 1y ago
I wondered about this, too. Cloudflare have some recent data about traffic from bots ( https://blog.cloudflare.com/from-googlebot-to-gptbot-whos-cr... ) which indicates that, for the time being, the overwhelming majority of t
17.
U.S. bombs Iranian nuclear sites(bbc.co.uk)
1247 points by mattcollins 1y ago|3848 comments
18.
How to Use AI in Software Product Development Today(mattcollins.net)
1 points by mattcollins 2y ago|0 comments
19.
PM plans to 'unleash AI' across UK to boost growth(bbc.co.uk)
3 points by mattcollins 2y ago|0 comments
20.
AI Helped Me Create Today's #1 Product Hunt Tool in Hours(mattcollins.net)
3 points by mattcollins 2y ago|0 comments
21.
99 Bottles of OOP now available in Python(sandimetz.com)
255 points by mattcollins 2y ago|86 comments
22.
Advanced Voice is now available in the ChatGPT app to Plus users in the UK(twitter.com)
1 points by mattcollins 2y ago|0 comments
23.
by mattcollins 2y ago
I noticed that, too. It does seem 'odd'.
24.
FastHTML: The Perfect Framework for Simple AI-Powered Web Apps?(mattcollins.net)
1 points by mattcollins 2y ago|0 comments
25.
Healthcare serial killer or coincidence? (2022) [pdf](rss.org.uk)
2 points by mattcollins 2y ago|0 comments
26.
What the Hollywood writers have agreed with the studios about AI
11 points by mattcollins 3y ago|3 comments
27.
Summary of the 2023 WGA MBA(wgacontract2023.org)
36 points by mattcollins 3y ago|51 comments
28.
by mattcollins 3y ago
Per the WGA's summary: 1) AI can’t write or rewrite literary material, and AI-generated material will not be considered source material under the MBA, meaning that AI-generated material can’t be used to undermine a writer’s credit or s
29.
by mattcollins 5y ago
Apply4 | Senior Ruby on Rails Engineer | Remote (EU timezone) | Full-time | $70k - $85k | https://bit.ly/ycror Our software helps local governments manage public outdoor spaces (parks, roads, etc.) more effectively, making
30.
Covid-19: Novavax vaccine shows 89% efficacy in UK trials(bbc.co.uk)
25 points by mattcollins 6y ago|12 comments

More ›