Posts by shieldsy05@aus.social
(DIR) Post #AZsfxLLxaqB1xz7eUa by shieldsy05@aus.social
2023-09-17T12:12:14Z
0 likes, 0 repeats
Okay, I’m WAY behind on my #Bird photos, so here’s another one!This is a #Tūi!Based on the date, I think this was taken at #LakeMatheson (I took a lot of bird photos there).#NewZealand #BirdPhotography #Birds #Nature #NaturePhotography #photography #BirdsOfMastodon
(DIR) Post #Al91iMdg5btnsoKNk0 by shieldsy05@aus.social
2024-08-20T02:40:26Z
0 likes, 0 repeats
@foone this is how I feel about computer programming and process efficiency in general. I’m not a programmer (i mean, I get by), but so many processes and systems in so many places where it’s just like “this process dumb and easily fixable with a small amount of simple code”.
(DIR) Post #AodLvK6OYB98Et5axU by shieldsy05@aus.social
2024-12-02T11:26:09Z
0 likes, 0 repeats
@foone yeah it tried to tell me that orthogonally adjacent meant diagonally adjacent and vice versa 😅
(DIR) Post #AuKlUdNwnHqQ1JaCYa by shieldsy05@aus.social
2025-05-21T01:51:27Z
0 likes, 0 repeats
Going to an AI talk for work today that’s supposed to cover AI trends and business readiness, and a “live copilot 365 demo”.Any questions I should prepare? :P
(DIR) Post #AuKlUeWqXgYbZCkp4C by shieldsy05@aus.social
2025-05-21T22:35:39Z
0 likes, 1 repeats
lol. If anything this talk just reinforced why we absolutely shouldn’t be using AI for our work.The first part of the talk was framed a bit like “is your business ready for AI”, and provided a lot of “questions you should ask yourself”. But didn’t provide any answers!!!Among which:- how will you validate AI generated data?- how will you monitor AI generated data?- who is accountable for the reliability of AI generated data?There was a lot of “trust, but verify”. Like, then what’s the point? Humans aren’t capable of this. Even if they start out verifying all the output, they’ll stop verifying very quickly, because it detracts from the return.
(DIR) Post #AuKlUm5URHQb0g15XM by shieldsy05@aus.social
2025-05-21T22:35:51Z
0 likes, 1 repeats
Other ‘key points’ made:- Data hygiene is crucial for AI- consider which of your companies data should be accessible to AI- use sensitivity labels and control listsOkay, in what industry are those things both in-place, and well managed????‘Oh and it’s important to note: if you’re using AI with sensitive data, make sure you have enterprise data protection, and that it’s switched on, otherwise you could accidentally be leaking data into the public AI dataset. Is that a data breach? Well, that’s up to your company to decide…! Make sure staff check that they’re logged into the web browser and they see the green shield here before putting anything confidential into Copilot chat’Have these people ever worked in a workplace? With other human beings!?
(DIR) Post #AuKlUttNQCUZDQPWlc by shieldsy05@aus.social
2025-05-21T22:38:38Z
0 likes, 0 repeats
The data they provided on AI ‘success’ was “users reported”. Which is gonna be biased as all hell.They also tried to say “oh, not using AI might lower staff retention, because staff want to use AI, and they don’t want to work for companies that don’t use AI.” Yeah, no.