lingo.lol is one of the many independent Mastodon servers you can use to participate in the fediverse.
A place for linguists, philologists, and other lovers of languages.

Server stats:

65
active users

#amazonqdeveloper

0 posts0 participants0 posts today
Paco Hope #resist<p>Here's a nice <a href="https://infosec.exchange/tags/darkpattern" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>darkpattern</span></a> from <a href="https://infosec.exchange/tags/AmazonQDeveloper" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>AmazonQDeveloper</span></a>. Installing it on my <a href="https://infosec.exchange/tags/debian" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>debian</span></a> laptop and it pops up this GUI. Now, grey text on white background is a completely pedestrian dark pattern.</p><p>The thing that is new for me, that isn't obvious from a single screenshot, is that the "skip" grey text doesn't appear at all for like 10-20 seconds. When you first see this dialog box, there's no 'skip' button. This might lead you to believe it isn't possible to skip. I tend to read things that pop up. And so while I was reading the skip button appeared.</p><p>Thanks <a href="https://infosec.exchange/tags/darkpatterns" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>darkpatterns</span></a></p>
Paco Hope #resist<p>More fun with <a href="https://infosec.exchange/tags/AI" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>AI</span></a> to start my day. A buddy of mine has a some web forum data where doctors and nurses answered people's questions: 20 years worth of Q&amp;A. He asks me how he might train an AI model on that data. I have a vague idea and tell him "you can fine tune a foundation model with that data." To save myself typing I ask <a href="https://infosec.exchange/tags/AmazonQDeveloper" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>AmazonQDeveloper</span></a> :</p><p><em>I have a bunch of posts on a web forum stored in a SQL database...I want to use Amazon Bedrock to create a chatbot that answers questions using a fine-tuned model. Write out the high-level steps I will need to take.</em><br>(I said a bunch of stuff in that ellipsis about what the data was)</p><p>Among the many steps it tells me to do, it says:</p><blockquote><p>Format your data according to <a href="https://infosec.exchange/tags/Anthropic" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>Anthropic</span></a>'s fine-tuning specifications for <a href="https://infosec.exchange/tags/Claude" class="mention hashtag" rel="nofollow noopener noreferrer" target="_blank">#<span>Claude</span></a></p></blockquote><p>That's interesting. So I ask: <em>where can i get more information on Anthropic's fine-tuning specifications for Claude?</em></p><p>It replies:</p><pre><code>Amazon Q is having trouble responding right now: <br> 0: unhandled error (InternalServerException)<br> 1: service error<br> 2: unhandled error (InternalServerException)<br></code></pre><p>I chuckle, press the up arrow, and resubmit my query. It replies:</p><blockquote><p>I apologize, but I need to correct my previous response. I made a mistake in suggesting that Claude could be fine-tuned through Amazon Bedrock. Currently, Claude models (including Claude Sonnet) on Amazon Bedrock are not available for fine-tuning by customers.</p></blockquote><p>Great. Thanks, Q. It's like working with a chipper intern.</p>