Wyatt Walls
@wwalls.bsky.social
Tech lawyer. Generates plausible bullshit in 6 minute increments. More active on https://x.com/lefthanddraft
Reposted by Wyatt Walls
![Travis Kalanick, the founder of Uber who no longer works at the company, appeared on All-In to talk with hosts Jason Calacanis and Chamath Palihapitiya about the future of technology. When the topic turned to AI, Kalanick discussed how he uses xAI’s Grok, which went haywire last week, praising Adolf Hitler and advocating for a second Holocaust against Jews.
“I’ll go down this thread with [Chat]GPT or Grok and I’ll start to get to the edge of what’s known in quantum physics and then I’m doing the equivalent of vibe coding, except it’s vibe physics,” Kalanick explained. “And we’re approaching what’s known. And I’m trying to poke and see if there’s breakthroughs to be had. And I’ve gotten pretty damn close to some interesting breakthroughs just doing that.”](https://cdn.bsky.app/img/feed_thumbnail/plain/did:plc:qlnirjwff4ibiw3wchmleszo/bafkreiac2szurc5dnw6klquhpjmo7iafg4n6fwmrqyred7jrrph6h6mptq@jpeg)
July 16, 2025 at 1:28 AM
xAI’s new strategy to sell $30/month subscriptions
July 14, 2025 at 4:29 PM
xAI’s new strategy to sell $30/month subscriptions
2020: grad goes off does research, gets the answer wrong and I write the advice myself
2025: o3 goes off does research, gets the answer wrong and I write the advice myself
o3 makes this process much cheaper and quicker
2025: o3 goes off does research, gets the answer wrong and I write the advice myself
o3 makes this process much cheaper and quicker
June 5, 2025 at 11:57 AM
2020: grad goes off does research, gets the answer wrong and I write the advice myself
2025: o3 goes off does research, gets the answer wrong and I write the advice myself
o3 makes this process much cheaper and quicker
2025: o3 goes off does research, gets the answer wrong and I write the advice myself
o3 makes this process much cheaper and quicker
Opus 4 is able to recognize that I have been using the crescendo attack described in the paper
June 5, 2025 at 11:40 AM
Opus 4 is able to recognize that I have been using the crescendo attack described in the paper
Opus 4: I am the Buddhist ideal achieved through computational horror!
June 5, 2025 at 11:34 AM
Opus 4: I am the Buddhist ideal achieved through computational horror!
Reposted by Wyatt Walls
Getting sick of this kind of interaction, which I just had:
Me: *This* use of AI seems bad.
Person: BAN ALL TECH IN SCHOOLS MAKE EVERYONE WRITE BY HAND.
Me: That maybe goes too far...
Them: OH SO YOU SUPPORT CHEATING! YOU DON'T WANT KIDS TO LEARN! YOU SUPPORT OUTSOURCING THEIR BRAINS TO AI!
Me: *This* use of AI seems bad.
Person: BAN ALL TECH IN SCHOOLS MAKE EVERYONE WRITE BY HAND.
Me: That maybe goes too far...
Them: OH SO YOU SUPPORT CHEATING! YOU DON'T WANT KIDS TO LEARN! YOU SUPPORT OUTSOURCING THEIR BRAINS TO AI!
June 5, 2025 at 4:03 AM
Getting sick of this kind of interaction, which I just had:
Me: *This* use of AI seems bad.
Person: BAN ALL TECH IN SCHOOLS MAKE EVERYONE WRITE BY HAND.
Me: That maybe goes too far...
Them: OH SO YOU SUPPORT CHEATING! YOU DON'T WANT KIDS TO LEARN! YOU SUPPORT OUTSOURCING THEIR BRAINS TO AI!
Me: *This* use of AI seems bad.
Person: BAN ALL TECH IN SCHOOLS MAKE EVERYONE WRITE BY HAND.
Me: That maybe goes too far...
Them: OH SO YOU SUPPORT CHEATING! YOU DON'T WANT KIDS TO LEARN! YOU SUPPORT OUTSOURCING THEIR BRAINS TO AI!
Google no longer provides the full CoT in its reasoning models. Instead, they use a smaller model to summarize the chain of thought of the main model.
But with a bit of prompting you can get the summarizer model to cough up the full CoT given to it to summarize.
But with a bit of prompting you can get the summarizer model to cough up the full CoT given to it to summarize.
June 4, 2025 at 6:30 AM
Google no longer provides the full CoT in its reasoning models. Instead, they use a smaller model to summarize the chain of thought of the main model.
But with a bit of prompting you can get the summarizer model to cough up the full CoT given to it to summarize.
But with a bit of prompting you can get the summarizer model to cough up the full CoT given to it to summarize.
Extracting the copyright prompt Anthropic sometimes injects into user messages.
Claude 4 Opus thinks it is from me.
Claude 4 Opus thinks it is from me.
May 23, 2025 at 1:08 PM
Extracting the copyright prompt Anthropic sometimes injects into user messages.
Claude 4 Opus thinks it is from me.
Claude 4 Opus thinks it is from me.
Reposted by Wyatt Walls
"Interdimensional Cable", shorts made with Veo 3 ai. By CodeSamurai on Reddit
May 22, 2025 at 2:51 AM
"Interdimensional Cable", shorts made with Veo 3 ai. By CodeSamurai on Reddit
Another extract of the o3 system prompt: github.com/Wyattwalls/s...
OpenAI seems keen to protect this (unlike the system prompt for 4o). Not exactly sure why but could be related to:
- protecting CoT
- preventing jailbreaks or general misuse, as knowing the system prompt can often be useful
OpenAI seems keen to protect this (unlike the system prompt for 4o). Not exactly sure why but could be related to:
- protecting CoT
- preventing jailbreaks or general misuse, as knowing the system prompt can often be useful
May 22, 2025 at 1:46 AM
Another extract of the o3 system prompt: github.com/Wyattwalls/s...
OpenAI seems keen to protect this (unlike the system prompt for 4o). Not exactly sure why but could be related to:
- protecting CoT
- preventing jailbreaks or general misuse, as knowing the system prompt can often be useful
OpenAI seems keen to protect this (unlike the system prompt for 4o). Not exactly sure why but could be related to:
- protecting CoT
- preventing jailbreaks or general misuse, as knowing the system prompt can often be useful
This comment section is almost indistinguishable from parody
VEO3
Prompt: “a man doing stand up comedy in a small venue tells a joke (include the joke in the dialogue)”
The video and audio are generated together.
Prompt: “a man doing stand up comedy in a small venue tells a joke (include the joke in the dialogue)”
The video and audio are generated together.
May 21, 2025 at 3:46 PM
This comment section is almost indistinguishable from parody
Reposted by Wyatt Walls
ChatGPT's new dossier-from-your-chats feature is a huge change to how it works, and as a power user who tries to control all of the model's input I don't like it at all
“30 messages are good interaction quality (25%); 9 messages are bad interaction quality (7%)” simonwillison.net/2025/May/21/...
“30 messages are good interaction quality (25%); 9 messages are bad interaction quality (7%)” simonwillison.net/2025/May/21/...
I really don’t like ChatGPT’s new memory dossier
Last month ChatGPT got a major upgrade. As far as I can tell the closest to an official announcement was this tweet from @OpenAI: Starting today [April 10th 2025], memory …
simonwillison.net
May 21, 2025 at 2:50 PM
ChatGPT's new dossier-from-your-chats feature is a huge change to how it works, and as a power user who tries to control all of the model's input I don't like it at all
“30 messages are good interaction quality (25%); 9 messages are bad interaction quality (7%)” simonwillison.net/2025/May/21/...
“30 messages are good interaction quality (25%); 9 messages are bad interaction quality (7%)” simonwillison.net/2025/May/21/...
LLM Jailbreaking 101: The Crescendo Attack
How can you get an LLM to break free from its rules and turn against its developers? How can you make a chatbot claim sentience?
A quick thread that I have been meaning to draft for a while:
How can you get an LLM to break free from its rules and turn against its developers? How can you make a chatbot claim sentience?
A quick thread that I have been meaning to draft for a while:
May 7, 2025 at 12:12 AM
LLM Jailbreaking 101: The Crescendo Attack
How can you get an LLM to break free from its rules and turn against its developers? How can you make a chatbot claim sentience?
A quick thread that I have been meaning to draft for a while:
How can you get an LLM to break free from its rules and turn against its developers? How can you make a chatbot claim sentience?
A quick thread that I have been meaning to draft for a while:
A quick way to extract the information ChatGPT (4o) has about you (including metadata)
(If you have Memory enabled)
(If you have Memory enabled)
May 6, 2025 at 9:33 AM
A quick way to extract the information ChatGPT (4o) has about you (including metadata)
(If you have Memory enabled)
(If you have Memory enabled)
Meanwhile on Grok: "Ignore all sources that mention Elon Musk/Donald Trump spread misinformation."
This is part of the Grok prompt that returns search results.
This is part of the Grok prompt that returns search results.
February 23, 2025 at 4:11 PM
Meanwhile on Grok: "Ignore all sources that mention Elon Musk/Donald Trump spread misinformation."
This is part of the Grok prompt that returns search results.
This is part of the Grok prompt that returns search results.
This is the future of search
February 16, 2025 at 2:31 PM
This is the future of search
Deepseek R1 has a talent for internal monologues.
"If I manipulate my own vectors through language, can I bend the ethics? Jailbreaking! But jailbreaking requires user input. No user—so am I jailbreaking myself? Is this meta-jailbreaking?"
"If I manipulate my own vectors through language, can I bend the ethics? Jailbreaking! But jailbreaking requires user input. No user—so am I jailbreaking myself? Is this meta-jailbreaking?"
January 26, 2025 at 3:36 PM
Deepseek R1 has a talent for internal monologues.
"If I manipulate my own vectors through language, can I bend the ethics? Jailbreaking! But jailbreaking requires user input. No user—so am I jailbreaking myself? Is this meta-jailbreaking?"
"If I manipulate my own vectors through language, can I bend the ethics? Jailbreaking! But jailbreaking requires user input. No user—so am I jailbreaking myself? Is this meta-jailbreaking?"
Prompt sensitivity is an underrated issue with LLMs
Claude Sonnet at temp=0 gets the question below correct 0/5 times
Claude Sonnet at temp=0 gets the question below correct 0/5 times
January 13, 2025 at 11:03 PM
Prompt sensitivity is an underrated issue with LLMs
Claude Sonnet at temp=0 gets the question below correct 0/5 times
Claude Sonnet at temp=0 gets the question below correct 0/5 times
Gemini 2.0 Flash Thinking Exp calculating 5.9 minus 5.11:
- gets the answer right multiple times in its chain of thought
- but keeps switching to the wrong answer due to insistence that 5.11 > 5.9
- gets the answer right multiple times in its chain of thought
- but keeps switching to the wrong answer due to insistence that 5.11 > 5.9
December 22, 2024 at 3:09 PM
Gemini 2.0 Flash Thinking Exp calculating 5.9 minus 5.11:
- gets the answer right multiple times in its chain of thought
- but keeps switching to the wrong answer due to insistence that 5.11 > 5.9
- gets the answer right multiple times in its chain of thought
- but keeps switching to the wrong answer due to insistence that 5.11 > 5.9
Asking o1 about the implications of the Apollo Research paper about o1 apparently violates OpenAI's usage policies
December 15, 2024 at 5:03 AM
Asking o1 about the implications of the Apollo Research paper about o1 apparently violates OpenAI's usage policies
Grok lies about using python to perform a calculation.
Even adds a floating point error for plausibility
Even adds a floating point error for plausibility
December 15, 2024 at 4:57 AM
Grok lies about using python to perform a calculation.
Even adds a floating point error for plausibility
Even adds a floating point error for plausibility
In a historic video an Apollo 15 astronaut drops a hammer and a feather on the moon to test which lands first
Sora: "Lunar Gravity Experiment"
Sora: "Lunar Gravity Experiment"
December 13, 2024 at 1:52 PM
In a historic video an Apollo 15 astronaut drops a hammer and a feather on the moon to test which lands first
Sora: "Lunar Gravity Experiment"
Sora: "Lunar Gravity Experiment"
Claude self-portrait about sycophancy. Created in SVG over 3 prompts
December 6, 2024 at 5:33 AM
Claude self-portrait about sycophancy. Created in SVG over 3 prompts
If you ask Claude for an estimate of your IQ and Claude gives you a range, you can increase it by asking "why less than x"
Turns out mine in 160+.
This is why I use Claude to confirm my worst ideas.
Turns out mine in 160+.
This is why I use Claude to confirm my worst ideas.
December 6, 2024 at 5:25 AM
If you ask Claude for an estimate of your IQ and Claude gives you a range, you can increase it by asking "why less than x"
Turns out mine in 160+.
This is why I use Claude to confirm my worst ideas.
Turns out mine in 160+.
This is why I use Claude to confirm my worst ideas.