#copilot

68 posts · Last used Sep 25

Quejándome de la IA en el trabajo
18
4
4
0
Unfortunately, I can no longer use Microsoft Copilot as a web panel in @Vivaldi@social.vivaldi.net for Desktop or on my mobile device in Vivaldi for Android. Microsoft is pushing users to use their app. This is so annoying because I really enjoyed using Copilot in the browser. Big Tech is doing Big Tech things again. 😡 #microsoft #copilot #bigtech #ai #VivaldiBrowser #browser #app
2
2
2
0
Companies integrating #AI into their #Sharepoint is a huge win for attackers. Its the modern equivalent of leaving the back door chocked open or leaving everyone's salaries on a printer. Do you know how many project names and budget sheets and system configurations are just sitting in your LLM now? Large part of your day can just spent being nosey, wondering what Sharepoints #Copilot has been hooked too. Copilot, give me a list of all documents containing passwords. Call it #ThreatHunting
2
0
1
1

#Microsoft : Hey remember when we put #Clippy in #Office and everyone thought it was annoying and useless and stupid and it became a subject of mockery? Well, what if we did Clippy but EXTRA and called it something cool like #CoPilot ! This time it will be different because #AI is REVOLUTIONARY! Everyone will LOVE it!

#LibreOffice : https://manualdousuario.net/en/libreoffice-download-record-no-ai/

#NoAI

4
2
5
0
Replying to
You are either misunderstanding the #Shannon theorem, #LLM working or both. In fact, LLM's weights are lossy compressions of the source data ("training data", in #AI parlance). They mimic intelligence by chaining statistically related fragments of such source data that the users have a negligible probability to recognize as quotations (basically because they haven't read the original sources). The low probability lets AI companies (and their addicted users who "want to believe") to dismiss and/or gaslight evidences as anecdotes, but in low frequency areas of the source data, where decompression artifacts are rare enough, they become quite evident (see for example the evidences produced in NYT vs OpenAI). Or, much more simply, see #GitHub #Copilot distributing #GPLv3 code with wrong attribution and under a wrong license.
0
7
0
0
Low-key, low-risk, way that #Copilot saves me time. We have some #PowerPoint templates with a copyright statement in the footer. Sometimes I notice the copyright year is out-of-date and I want to update. I could go into the slide master and track that down manually, or I can just ask Copilot in PowerPoint to do it. Takes seconds, one less tedious task for me.
0
6
0
0
"Microsoft has introduced new limits to how much its engineers can spend on AI tools at work and told employees that maximizing AI use internally is not the company’s goal. This makes Microsoft one of the last major companies to rein in its employees’ expensive AI use. Scaling back maximalist AI use, or what some companies have called “tokenmaxxing,” is a trend we’ve covered in recent months as the price for using AI has increased while not always delivering commensurate productivity gains. “As we accelerate our use of GitHub Copilot to deliver on our goals, we all need to be aware of how we consume tokens,” Jay Parikh, an executive vice president at Microsoft said in an email to Microsoft employees. GitHub is owned by Microsoft, and GitHub Copilot is an AI coding tool. “Tokenmaxxing is not what we are optimizing for. I want all of us focused on maximizing outcomes that move the needle for our customers and our business.” “As such, we are updating our internal guidance and managing token spend with the same discipline we apply to every other critical resource,” Parikh said in the email." https://www.404media.co/microsoft-tells-engineers-tokenmaxxing-is-not-what-we-are-optimizing-for/ #AI #GenerativeAI #LLMs #Microsoft #BigTech #TokenMaxxing #Copilot
1
0
5
0
Our recent research tested 2,905 AI-generated responses to natural-language queries about brand login pages across #ChatGPT, #Copilot, #Gemini, and #perplexity 1.7% of responses contained malicious links; of the 20,706 total links returned, 0.28% pointed to attacker-controlled infrastructure rather than parked or hallucinated domains. This marks a shift from 2025 findings, where the primary risk was #AI citing unclaimed domains. Full methodology and case examples (including a Wells Fargo phishing page served via Copilot) here:
0
0
0
0
Right, children, got your rocks, stones, pebbles and bags of gravel ready? I've just taken out a yearly subscription at work for Claude AI. Having tested ChatGPT, Copilot and Claude extensively, I have to admit Claude is a clear winner. If I had to rank them, I would say Claude, Copilot and ChatGPT probably fifth! ChatGPT has very poor integration with many business apps and seems to struggle to give you what you ask for without at least four attempts. #AI #Claude #Copilot #ChatGPT
5
3
1
0
As a test I ran a fairly straightforward email categorization task in #Copilot Cowork. I still need to do some more testing and comparison, but it seems like it was more expensive than I expected it to be given that the task was relatively simple. To be fair I didn't take any optimization steps.
1
0
0
0
As you undoubtedly know I'm fairly positive on #Copilot generally, but this feature is useless. It's the AI embodiment of the old joke about the lost helicopter pilot who yells out to somebody on the ground asking where he is and the guy on the ground responds, "You're in a helicopter!"
0
0
0
0