@bdambrosio I dunno, it just looks like Mythos crossed a threshold of usefulness for cyber capabilities. That doesn't require a magical step change, capability gain can just cross thresholds of utility as part of standard progress. https://t.co/yfhz7ignmO

After seeing that Claude Mythos marketing turned out to be, as expected, a scam, I wanted to make a master list of tricks being used to market LLMs. The master list includes statements directly from leadership in the companies or from the "organic marketing" of people on social media, along with an explanation on how the scam works. This is my first attempt, so likely incomplete. The LLM Marketing Scams Master List v1: "Two more weeks" - the models will be good enough someday soon to do what we claim. "They're already good enough" - the models are already good enough to replace workers, but it hasn't happened yet because of x y z reasons. "We just built God in the backroom, and no, you can't see it" - the models they built in private are actually capable of doing the things we have been waiting for, but they can't let us see them yet for x y z reasons. "Actually they already have replaced jobs" - the layoffs that tech companies have been doing, citing AI as the reason, have already been replaced with current LLM tech, ignoring market conditions and past data on layoffs during such conditions. "You just don't know how to use then as well as me" - the models are good enough, but esoteric prompt engineering is required to get these results, and no, I won't teach you. "I built an app making big money with LLMs" - they claim they already have made startup companies, almost always SaaS companies, that are making them tons of money, but when you ask to see them, they won't show you. "You aren't using the right model" - claims that you must be using the wrong model and need to use Open Claude 420b-parameter Gemini Plus Pro 6.9 with 4RealThisTime HomerSimpson agent mode enabled. Note that this will be used to attack every study on the effectiveness of LLMs, since studies take time to complete and publish, with new models releasing more frequently than it's possible to complete and publish a study "You're falling behind" - claims that you need to use the bots now, even though they aren't good enough to fully automate any jobs, because otherwise, when the bots are good enough, you will lose your natural English skills required to prompt effectively. "All these companies are using LLMs, so do you think you know better than they do?" - pointing to claims of large companies deeply invested in LLMs being a success saying that LLMs are being used effectively, with no viewable results in the speed and/or quality of their company's output. "The benchmark score went up" - claiming improvements on the benchmarking tests given to their latest model, despite the training being specifically tuned to improve on these tests, and then conflating better benchmark scores with actually being more able to automate jobs or drastically improve worker productivity. "It can now count the letters in Strawberry/can now do things it famously couldn't do previously" - saying that it can now count the letters in Strawberry or instruct you on how to use a cup without a bottom, etc. is often done to suggest increased reasoning for the LLM, but often involves just hard coding an answer into the service. "It has escaped our control" - saying that they cannot control the LLM, implying it is conscious or living to some degree when really it just said words that it wasn't supposed to or an agent used an app that wasn't intended by the user's prompt when next-token predicting "It's feeling sad/scared/happy/angry, suggesting it is conscious" - they ask the LLM how it is feeling, and it next-token predicts a response that includes an emotion felt by humans, since training data is from human conversations online. "Costs are going down/the LLM service is profitable" - ignores training costs and capex for hardware, usually just referring to inference being profitable, which isn't even true in many cases. Training and capex is 95%+ of the total costs to serve the models. Did I miss any?

Elon Musk: “You’d think that if you’re going to be appointed as a judge, you have to have proven that you have an excellent knowledge of the law and that you will make your decisions according to the law. That’s what we assume should be a requirement. You’ve got to know what the law is, and then you need to make decisions in accordance with the law. Not vibes. You can’t be just vibing as a judge.”

🇿🇦Each cross represents a white farmer who was murdered in South Africa. And some people still deny that white South Africans/Boers are persecuted and say they don’t deserve asylum because they are white. This looks like a genocide. https://t.co/5JCMEKkToc

This is the best arguments I have heard against moon landing hoax. I love this. https://t.co/PlrOqS4R2W

This is the best arguments I have heard against moon landing hoax. I love this. https://t.co/PlrOqS4R2W

When President Trump called out the racism against white farmers and their families in South Africa. https://t.co/ePGOB6rvHq

@nikitabier https://t.co/rrpgembZyN

I think it needs a name https://t.co/1sf6kYl9vy

@alanhoward Yeah cause everyone just uses the ForYou feed and takes what the algorithm shows them. I built my own algorithm: https://t.co/kiuZ7QXLzb

envy drives so much of human behavior people who are smart and envious are especially dangerous because they will rationalize their envy until they feel *virtuous* in stealing from and killing those they envy https://t.co/VyrUTzJhiv

NEWS: UK carrier Virgin Atlantic has brought forward its plans to equip its aircraft with @Starlink Wi-Fi. The original plan was to install Starlink in Q3 of this year, but the revised plan will see the airline's first aircraft with Starlink in May (next month). VA says the first of its aircraft to be equipped with Starlink will be an Airbus A350-1000 aircraft, and the first flight equipped with the system will operate between London Heathrow Airport and New York JFK Airport. Starlink can be installed in a fraction of the time it takes other internet systems to be installed, so the rollout for other aircraft should be relatively quick.

You don’t understand how comfortable the Cybertruck is until you try it. I highly recommend you take a test drive. Its ride quality is better than the Cadillac Escalade. https://t.co/5kjlhwI7Dq

@chatgpt21 Which is why I built https://t.co/kiuZ7QXLzb My AI reads tens of thousands of posts and finds the good stuff in the AI community

@recatm Xie Saining is spot on. Internet text comes pre-baked with human POV and labels. Raw images/video don’t. One scene can be shot from endless angles, camera height, distance, tilt, 1st-person vs 3rd-person, each completely changing the signal. This is why vision needs far stronger inductive biases than language. Text has it embedded. Vision is missing data to fill this gap.

wtf. This is the atmosphere certain people and orgs have directly / indirectly caused. https://t.co/wnGqe2J2u9

@LexnLin https://t.co/WaavXl0Ne6

fwiw we have already helped save thousands of lives, very very indirectly. we are aware of at least one large frontier lab that has benefitted from our benchmark suite to improve their own internal medical benchmarks. Btw if other frontier labs want to improve their medical benchmarks we'd love to chat!

Super excited to Introduce our latest work: Squeeze Evolve. We unify test-time scaling methods into one evolutionary framework — then orchestrate many models across it. 3x lower cost. 10x throughput. 97.5%(SoTA) on ARC-AGI-V2. No verifier required. Framework: https://t.co/5hmOyZvSKU

@ChiefAgenteer Use https://t.co/kiuZ7QXLzb then. It finds the good stuff out of 50,000 accounts

the vibes at coachella were great thanks beibs! https://t.co/A7RD50BZUu

MiniMax M2.7 is now on Together AI. Trained by letting it run its own RL loop, resulting in the highest open-source score on MLE Bench Lite. https://t.co/3E5mh9EDdz

Is X dying? My feed is dramatically better than it was months ago. And my lists are busier. I went to threads and it just isn’t as good. And I can’t make this on threads: https://t.co/8L5xphk0qQ So I am not leaving. And yes reach is down. Why? Grok is getting pickier. Which is why my feed is better. Which pisses off people because they used to get more reach. Me included. But my feed is better. Which tells me the service is improving. Which tells me that soon the numbers will start going up. Which is why I am not leaving.

https://t.co/V0J2P62KSp https://t.co/dErBykXPPg

5 years ago, my little girl was secretly being groomed and transitioned at her public school. 4 years ago, she was still lost in a dark cloud, suicidal, identifying as a “boy” named Toby. Last night, she went to her first prom with her boyfriend — glowing with confidence and grace. Thank you, Jesus, for saving her. 🙏🏼 What the enemy meant for evil, God is turning into good. If you’re a parent fighting for your child right now and it feels hopeless, please hold on. There is light on the other side. I won’t stop fighting for your horror story to have a happy ending, too.

“A kid these days could start a multi billion dollar business by using these tools in some way no one has thought about” Enroll and give a shot at securing $1m in funding! https://t.co/6WQrkloa9l

“There's a lot of narcissism in self-hatred.” — David Foster Wallace https://t.co/YpfISKLVPW

“There's a lot of narcissism in self-hatred.” — David Foster Wallace https://t.co/YpfISKLVPW

@paularambles wait did you see us? I was dressed like https://t.co/Qa2NoNIStM

Looking forward to CHI this week! We have a ✨Best Paper ✨ on a "what-if" analysis tool for RAG. Reach out to chat! I'm interested in: MLOps/LLMOps, data analysis, and better interfaces for human-AI collaboration (and, very soon i'll be recruiting students/postdocs) https://t.co/JVeiFbTOHM

https://t.co/dZiZ5HVgIl

check out the book of disquiet https://t.co/Q2SxupbH1N