The blog post recounts the author's experiences at BABSCon 2025, the final convention for My Little Pony fans, known as bronies. It discusses the evolution of the brony community, the decline of conventions, and the author's perso...
About:
Alex Irpan is a Senior Research Scientist at Google DeepMind with interests in AI, robotics, and enjoys card games and puzzlehunts.
Website:
Specializations:
Interests:
Incoming Links:
Outgoing Links:
Subscribe to RSS:
A personal account of the author's experience at MIT Mystery Hunt 2026, detailing team dynamics, puzzle-solving strategies, and reflections on the event's challenges and structure.
The author supports Anthropic's lawsuit against the Department of War, criticizing government overreach in AI military contracts and emphasizing the need to protect civil liberties.
The post discusses the evolution and perception of Auto-Tune and generative AI in music and art. It highlights how Auto-Tune, initially a tool for pitch correction, became a controversial symbol of inauthenticity in music, particu...
4Ten Years Later
2025-08-18 • blogging writing creativity video games
The author reflects on their ten-year blogging journey, expressing gratitude for their readers and discussing their motivations for writing. They share insights about their writing process, the challenges of maintaining a consiste...
0Who is AI For?
2025-04-01 • programming technology social sciences ethics
The text discusses the current uses of AI, focusing on its applications in image generation, coding assistance, speech-to-text, and text-to-speech. It also delves into the focus on coding in AI development, the limitations of AI i...
0MIT Mystery Hunt 2025
2025-01-28 •
The text discusses the MIT Mystery Hunt 2025, including the author's experience, the puzzles, and the overall success of the event. The author compares this year's hunt to previous years and shares insights into the puzzles and me...
0Using AI to Get the Neopets Destruct-o-Match Avatar
2025-01-10 •
The text discusses the author's experience with Neopets and the pursuit of avatars, particularly the Destruct-o-Match avatar. The author explains the appeal of Neopets, the difficulty of obtaining avatars, and the specific challen...
0Late Takes on OpenAI o1
2024-12-04 •
The text discusses OpenAI o1, a model release believed to be a post-trained version of GPT-4o. It is trained to spend more time generating and exploring long, internal chains of thought before responding, especially for domains li...
0Nine Years Later
2024-08-18 •
The author reflects on the past year, discussing their break from puzzle writing, career change to AI safety, and new hobbies such as video games. They review various video games they played, share statistics on their blog posts, ...
0I'm Switching Into AI Safety
2024-08-06 •
The author discusses their decision to switch from working on robotics projects to joining the AI safety team at Google DeepMind. They explain their reasons for the switch, including a desire for change, curiosity, and the potenti...
0The Tragedies of Reality Are Coming for You
2024-07-08 •
The text discusses the challenges of robotics and machine learning in dealing with the complexities of reality. It highlights the difficulties in bridging the gap between the physical world and software, and the limitations of cur...
0Puzzlehunting 201
2024-04-30 •
The author is an experienced puzzlehunter who has won several puzzlehunts and has been doing hunts for over a decade. They have decided to write about the habits they have learned along the way. The post is not for beginners and a...
0Solving Crew Battle Strategy With Math
2024-03-24 •
The text discusses the crew battle strategy in Super Smash Bros tournaments, focusing on the optimal strategy for sending players into 1v1 matches. It delves into the mathematical aspects of player matchups and probabilities, conc...
0MIT Mystery Hunt 2024
2024-01-21 •
The text is about the author's experience at the MIT Mystery Hunt 2024. They talk about their pre-hunt activities, the challenges they faced during the hunt, and their thoughts on the puzzles and the overall experience.
0My AI Timelines Have Sped Up (Again)
2024-01-11 •
The author discusses the changes in their AI timelines, comparing their 2020 forecast to their current one. They talk about the role of compute, unsupervised learning, transfer learning, and scaling laws. They also discuss the ris...