Rachel Appel
LightNews LightNews
banner
rachelappel.bsky.social
Rachel Appel
@rachelappel.bsky.social
1.3K followers 310 following 320 posts
.NET & Rider developer 🥑 at @JetBrains • shenanigans🍀• shitposts 💩 • MSc Software Engineering 🖥️ • 🧙‍♀️
Posts Replies Media Videos
rachelappel.bsky.social
A carton of a hallway with 3 doors. The first says Azure, is open and blood is flowing out of it. The second says AWS, is open and blood is flowing out of it. The 3rd says Cloudflare and shows the Grim Reaper knocking on it.
November 18, 2025 at 9:55 PM
rachelappel.bsky.social
I'm amused. Exclamation mark.
Two-line error message. First line says "Sorry, something went wrong" the next line says "Exclamation mark"
November 18, 2025 at 1:58 PM
rachelappel.bsky.social
Top frame: A hand holding a poisonous mushroom and a stick figure labeled "AI" is looking at it. Caption: "Is this mushroom edible?" Answer from AI: "Yes" Bottom frame: A man in a hospital bed, with the AI stick figure next to him. Caption "You're right, it's a poisonous mushroom. Would you like to know more about poisonous mushrooms?"
October 25, 2025 at 6:28 PM
rachelappel.bsky.social
Text: 4 out of 5 people enjoy code reviews. Image is of 4 people in Roman style attire beating one person with sticks.
September 25, 2025 at 10:39 AM
rachelappel.bsky.social
"In this book I will not attempt to answer the question of whether or not computers can be programmed to think. I sometimes ask people if people can think. Human intelligence may also be an illusion.”

-Timothy J. O'Malley
A book cover that says "ARTIFICIAL INTELLIGENCE PROJECTS FOR THE COMMODORE 64." Author: Timothy J. O'Malley On the cover is an abstract picture of a human head with lines going through it to represent AI, and below the head is a Commodore 64 keyboard.
September 15, 2025 at 11:16 AM
rachelappel.bsky.social
Ever wonder what happens when you send the wrong file to that AI recruiter at Meta? 😂
An email conversation between a candidate and an AI recruiter. The candidate (inadvertently?) sent a dick pic, and the AI "appreciates his confidence" and replied with it must be "meant to spart a conversation, I'm game, what's the story behind it?" 🤦‍♀️
September 4, 2025 at 9:33 AM
rachelappel.bsky.social
10 REM WORKING ON MY TALK
20 FINISH SLIDES
50 FINISH DEMOS
100 GIVE TALK
September 2, 2025 at 10:24 AM
rachelappel.bsky.social
Here’s a JPEG of a JPEG.
Graffiti that says “JPEG”
June 23, 2025 at 10:53 AM
rachelappel.bsky.social
A Buddah with monks sitting around him. One monk asks the Buddah "What makes us human". The reply is "Selecting all images with traffic lights".
June 8, 2025 at 2:21 PM
rachelappel.bsky.social
If the Muppets were real people...
Series of pictures of what the Muppets would look like as real people.
June 2, 2025 at 3:42 PM
rachelappel.bsky.social
Image of Nosferatu in a coffin. Caption says "when you've been up all night coding, and you start to hear birds chirping"
May 19, 2025 at 8:07 AM
rachelappel.bsky.social
😂
A fake book called "Vibe Coding" and the subtext says "I'm a developer now" and it shows Ralph Wiggums from the Simpsons on the cover.
April 24, 2025 at 1:42 PM
rachelappel.bsky.social
A classic
Drawing of an office with a bat in the CEO's chair. The caption says "autoexec.bat"
April 15, 2025 at 8:21 AM
rachelappel.bsky.social
😂😂😂
3 panels. 1st and 2nd panel show a bug teaching other bugs math, in which they all get the calculations wrong (for example 2+4=7). 3rd panel the teacher says "Great! time to destroy the programmers"
April 9, 2025 at 11:08 AM
rachelappel.bsky.social
No doubt whoever made this did it just after taping the remote so that the only buttons are volume up/down, channel up/down and on/off.
pic of a computer desktop with a bunch of characters from various memes all surrounding and pointing at the Chrome icon, which is in the middle of the screen.
February 11, 2025 at 4:04 PM
rachelappel.bsky.social
Current status
A nurse's office in a school, but it's a school for planets. The nurse is bringing in Earth, and the other planets are teasing Earth singing "Earth has people! 🎶" The nurse says "Now now, children"
January 26, 2025 at 4:47 PM
rachelappel.bsky.social
"AI is taking my job"
But not today. Today it's on vacation.
Breaking news that says "Chat GPT is down, website is offline"
January 24, 2025 at 5:20 PM
rachelappel.bsky.social
Every day
.
January 23, 2025 at 1:27 PM
rachelappel.bsky.social
😂
caption: "meal prepping has changed my life. i cannot recommend it enough" pic: large ziplock bags holding liquor and kitkats, looking like meal preps.
January 18, 2025 at 11:19 AM
rachelappel.bsky.social
Go home LinkedIn, you're drunk. 🍸🍺

After turning off notifications on a post, I still get notifications but it now also says "you turned off notifications about this post" FFS 🤦‍♀️🤦‍♀️
LinkedIn notification saying you turned off notifications about this post
January 10, 2025 at 9:51 AM
rachelappel.bsky.social
An AI site is asking to prove I'm human and not an AI. 🤦‍♀️🤣
The site "there's an ai for that" (theresanaiforthat.com) is asking you to verify that you're human.
January 8, 2025 at 12:20 PM
rachelappel.bsky.social
😆
Left panel: Christ the Redeemer statue Right panel: A wine bottle with a bottle opener, halfway opening it with its "arms" out like the original statue and a similar background of mountains and valleys.
January 7, 2025 at 7:56 AM
rachelappel.bsky.social
Stimmt das 😅
January 6, 2025 at 10:50 AM
rachelappel.bsky.social
Happy new year! 🎉
Sticker from Best Buy that says "Remember to turn off your computer before midnight on 12/31/99"
January 1, 2025 at 9:26 AM
rachelappel.bsky.social
Amusing myself today with gems like this.
"Why do LLMs make such long-winded responses?"
AI prompt "Why do LLMs make such long-winded responses?" AI reply "Large Language Models (LLMs) like me often generate long-winded responses for several reasons, usually rooted in how they are designed and trained: 1. **Objective During Training**: LLMs are typically trained to predict the next word in a sequence based on the context provided. When asked a question, the model aims to provide a thorough and contextually appropriate response to cover all possible nuances or interpretations. This can lead to overly detailed or verbose explanations. 2. **Uncertainty About User Intent**: LLMs cannot directly infer exactly what you're looking for unless the question is highly specific. As a result, responses may include elaborations, clarifications, or even tangentially related information to maximize the likelihood of satisfying the request. 3. **Training on Diverse Data**: LLMs are trained on a wide range of texts, from academic papers to casual articles, many of which contain elaborate and detailed discussions. This exposure influences the model's tendency to mimic the verbosity of its training data. 4. **Politeness and Accommodation**: Many LLMs are fine-tuned to align with user preferences, including being polite and helpful. This fine-tuning often leads to the model erring on the side of being overly explanatory to seem more friendly and accommodating, especially when unsure how concise the user prefers the response to be. 5. **Avoiding Ambiguity or Incompleteness**: By providing detailed responses, LLMs aim to reduce the chance of omitting important information or appearing unhelpful. They may overcompensate to ensure that the response is authoritative, even though it might seem excessive at times. 6. **Repetition and Redundancy**: LLMs don't have intrinsic awareness of when they've already sufficiently answered a query. Without explicit instruction to " -> there is more but it goes past the 2000 character limit
December 17, 2024 at 1:15 PM