🤖📈 Autonomous Agents Are The Fastest Growing Open-Source Technology In History - But... what are autonomous agents? - How do they work? - How can you build or use one? I wrote you the best overview on the planet. Get ready for your mind to explode 🤯 mattprd.com/p/the-complete-b…

Coding is 1% writing code and 99% figuring out why it doesn’t work.

More than 1 million people are using ChatGPT. But 99% of people aren't using it right. Here's how to transform your daily workflow with AI: (you can use these prompts right now👇)

This is an INSANE use of AI. I love it. All forms of media are going to be majorly transformed with AI in ways we can't yet predict. - Movies / TV - Video games - Music This is just the beginning. via @Flawlessai

I follow a lot of programmers on Twitter who are experimenting with AI. Every day they are releasing new AI products they build over the weekend. Here are 8 AI experiments that caught my eye this week:

In the future you will do one of the following: 1) Work FOR an AI 2) Work WITH an AI 3) Lose your job because you refused to do 1 or 2 If you refuse to work with AI, your job will be eliminated.

📽️ I just used AI to make a video explaining generative AI. What is it? How will it be used? AI script written by @OpenAI Person in video generated by @synthesiaIO Background b-roll created by @pictoryAI Combined everything with Final Cut Pro Took a few minutes total!

AI founders making AI apps by connecting 5 other AI apps together

When I was growing up we got: - Social media (Myspace, Facebook, etc) - Unlimited emails (gmail) - Search everything (google) - Really tiny computers (iPhones) Kids growing up today got: - Magic money (crypto) - Magic computers (AI) - Magic worlds (VR) What will my kid get?!

Wow! Google just released an AI that creates music. Type a description of a song and the AI makes it. The quality is really high 🎶😍 Listen google-research.github.io/se… >>> First you could generate text >>> Then images >>>Now music >>> Soon video What will AI be able to do next?

At this rate the GPT 5 demo will probably cure cancer

The world is about to change at an uncomfortable speed. Old: Programmers build with computers in computer language (code) New: AI let's you build with computers in human language More people making software => 📈 software => 📈 competition => VERY rapid 📈 in capabilities

🖼️➡️📽️ I just used AI to animate an image. >>> I took an image of the ocean >>> Uploaded it to @genmoai >>> It was automatically animated It took 30 seconds total!

Here are 9 fascinating AI tweets that caught my attention this week. 👇

And 1000% begging ChatGPT to fix it over and over

I love this! This is fantastic placement in Google search for Curlsmith's @OctaneAI quiz. Add a quiz to your @Shopify store = talk to every curious customer who visits and instantly personalize their shopping experience = $$$ Try out there quiz here: curlsmith.com/pages/quiz

AI is going to be everywhere in 2023. But not in the ways that you think. Here's how I predict AI will be incorporated into your daily life over the next 12 months: (some of these are already happening 👇)

AI is going to transform content creation as we know it. Instead of costing millions of dollars to make a movie or show... It will cost next to nothing. You think there are a lot of TV shows now? It's going to 1000x because of AI. 🤯 Here are 3 mind blowing examples:

Oh, just the CEO of $100mil+ in revenue @jonesroadbeauty saying the new AI product from @OctaneAI is "Game changing" for commerce brands. Probably nothing.

🗨️ Ask ChatGPT to help you create Twitter content that’s more likely to be shared. Feed it the text you want rewritten, and ask it to maximize the chance of it going viral.

Want to learn about AI? 🧠 Follow these AI pioneers to fill your feed with daily inspiration: @karenxcheng @gdb @alexandr_wang @deviparikh @alexgraveley @JackSoslow @DrJimFan @venturetwins @javilopen @realSharonZhou @dannypostma You're welcome ❤️

📈 Companies are grown by communities. 🚫 But the communities don't have any ownership in the company. This is wrong. We're doing things differently. Introducing the @OctaneAI Collective. We are making 50+ community members co-owners of @OctaneAI. Here's how it works 🧵👇

I've never worked at @Hulu. But in 2007, when I was 18, I launched Hulu publicly from my bedroom. Their team was completely caught off guard. Here's what happened 👇

Prediction: There are over 25 million ecom brands. The ecommerce teams that jump into AI first are going to become the BIGGEST money makers. Here's why. AI can: - Do tedious work for you - Replace people on your team and reduce the number of new people you need to hire - Analyze larger data sets more consistently than a person can - Handle most customer support inquiries - Handle most sales conversations - Personalize product recommendations and marketing content - Generate product and marketing images - Assist with strategy - Build websites and digital experiences Soon you will have AIs that you don't even have to assign individual tasks to, you'll just be able to say "your job is to optimize the website", and the AI will just... do it. Constantly. Like a person. But better. And AI is so cheap compared to people. Ecommerce brands will be able to operate as if they have hundreds, thousands, or millions of smart people but at the fraction of the cost. This may sound unbelievable, you can't even imagine millions of people working on your brand, what would they even do? You will have an entire team of AIs dedicated to every single customer. Imagine a customer of a skincare brand having these AIs, each one just for them: 1. Personal Skincare Advisor AI 2. New Arrivals Curator AI 3. Customer Support AI 4. Review and Feedback AI 5. Skin Health Tracker AI 6. Appointment Scheduler AI 7. Loyalty and Rewards AI 8. Skin Problem Solver AI 9. Routine Reminder AI 10. Virtual Try-On AI It's a completely new paradigm. First we're going to mirror our human org charts with AIs because we're used to that, then we're going to modify the org chart to allow for many more AIs, and then we will finally let AI run the org chart and it will be something we've never seen before. Not only are your customers going to have a team of AIs, you and each of your team members will have a team of AIs. Here's the kicker though, if you want your business to grow faster than everyone else's because of AI you have to dive headfirst in NOW before everyone else is doing it. You need to be ahead. What do you think happens to the first teams who figure out how to scale their business with AI? They are going to dominate their respective industries. So, where do you start? How do you keep on top of it all? Last night I made a resource just for you: aicheckout.io This is where you will: - Keep up with the latest AI developments and how they impact your ecommerce business - See behind the scenes at how real ecommerce operators are using AI today - Learn how to implement AI, ChatGPT, LLMs, media generation, etc into your workflows - Learn to surround your team and customers with value add AIs I've been building AI for almost 10 years now, I am coding AI agents and assistants every day for my own team, and my team and I actively support over 3,000 brands with AI implementations. I spend ALL of my time on this and I'm going to share everything with you. Conclusion: We live in an exciting world where things we didn't think were possible are being made available to us every day. It's up to us to be early adopters and run with it.

Two types of people on my feed: 1. People not building AI: “AI is just hype, just like web3, vr, this will die out, ignore it.” 2. Anyone who has even built the teeniest of things with AI: “Oh WTF.” Which one are you?

The upcoming launch of @Apple’s AR glasses, combined with the sudden acceleration of generative AI, may disrupt the tech landscape more than anything that’s ever happened before. Bigger than launch of: - Web browsers - iPhone - Social networks Huge opportunity for builders.

AI will fundamentally change how digital products are designed. Forever. The most exciting part of this? If you just sit and think you can likely come up with a way to use AI that nobody has ever done before. Try it.

This is by far the most impressive part of @OpenAI this weekend. No matter what is going on, customers are being treated as #1. ❤️

I don't think anyone can actually predict where AI will be in 5 years, the pace of development is beyond prediction.

Sam Altman doesn’t have equity, he has loyalty

Prediction: 6 months until there is a $1,000/month ChatGPT plan

Google

I love this! Bambu Earth spends $350,000/month on ads to drive traffic to their @OctaneAI quiz, making over $10M/year. Add a quiz to your Shopify store to guide every visitor and personalize their shopping experience = $$$ See exactly how they do it: octaneai.com/blog/the-bambu-…

/Example prompt/ Write a blog post about the importance of developing a morning routine that is: - no longer than 500 words - uses no more than 5 examples - speaks to entrepreneurs and creatives

We're building something really exciting into @OctaneAI & every day I have to stop myself from sharing it too early. It's magic ✨ Collective members will get first look Once they've used it for a bit, & give feedback, then we will start selectively sharing it w/ more people.

P.p.s. If you don’t know how to code I encourage you to just dive in ❤️

This is a first, and really wild. Someone added me on LinkedIn and messaged me saying they asked ChatGPT for help with chatbots and it told them to talk to me. 🤯

Our entire team is now harnessing the power of AI through GitHub Copilot and ChatGPT PRO. Rather than waiting for individual adoption, we're conducting a company-wide rollout & training to equip every team member with the skills to cowork with AI and increase their abilities.

The stages of building with @OpenAI 1) Holy shit 2) What’s a completion 3 Wait, tokens aren’t words? 4) I wish I could use more tokens 5) How does long-term memory work 6) How many #s are in an embedding vector?! 7) What are autonomous agents? 8) Holy shit What stage are you?

/Example prompt/ Rewrite the following tweet to be more likely to go viral: "Copywriting is an essential skill for entrepreneurs in 2021!"

Ok! Let's try an experiment. Meet @yesnoerror, a crypto AI agent that uses @OpenAI's o1 to find errors in published research papers. The AI agent uses it's $YNE coin to fund it's AI review system. The agent's wallet is starting with enough to run for 2 days.

Introducing GoalPillars - Put in your goal and get your own 64-cell Harada Method inspired dream sheet in seconds. goalpillars.com Very cool way to map goals.

Napster, 1999: Steal their song AI, 2023: Steal their voice

🗨️ Give it constraints If you give ChatGPT a specific task with constraints, it can come up with creative solutions and examples.

Excited to delve in

AI is going to completely change how brands operate in 2023. Here's a look into what this year holds 🧠✨ Contributions from @lovevalgeisler @eliweisss @Conniesends @babb_sophia @jsappington @Carson @samarowais @heykatgarcia @andrewjfaris & more octaneai.com/blog/how-brands…

AI is the new electricity. It will disrupt many industries and create new opportunity for the ones that embrace its power. Here are 7 AI products and experiments to inspire you to "think in AI" 🧵👇

Next up for @yesnoerror $YNE: ✅ Reviewing research papers every 10-60min ✅ Website stable with all the traffic ✅ AI agent now has ability to post review results on X ✅ Spoke with top AI researchers and mapped out how to increase accuracy of review method ⬜️ Research all existing methodologies used for research review (meta-analysis, etc) ⬜️ Implement more stable system for running reviews ⬜️ Add error checks incase the agent fails at pulling data from a paper ⬜️ Pagination for paper reviews (vs one long feed) ⬜️ Ability to filter paper reviews by errors / discrepancies ⬜️ Create synthetic data set of research papers with artificially inserted errors ⬜️ Improve use of @openai's o1 model to detect errors in synthetic data ⬜️ Release aggregate results across large number of papers This is phase 1 of a much larger project. This will not all happen overnight. Let's get it right. Let's audit all of science.

I'm thrilled to introduce @OctaneAI's latest innovation, Quiz AI, powered by @OpenAI! Quiz AI is a personal shopping assistant that any @Shopify store can easily integrate into their website in just minutes. Here's how it works:

A quiz that generates over $10mil/year? That's exactly what @BambuEarth is doing with their @OctaneAI quiz. It's deeply integrated with their tech stack: @Shopify @Klaviyo @Recartapp Here's how their strategy works:

Quizzes are THE BEST WAY to increase conversions on your @Shopify store 🔥 - Get to know customers in a fun way - Auto recommend personalized suggestions - Use data to personalize email/sms marketing Here are 8 absolutely STUNNING quizzes built with @OctaneAI

Brands that go ALL IN on AI are winning. You can literally make the customer feel as if you are paying 100% attention to them, because with AI you ACTUALLY ARE. I love this quote that @Bill_Gross shared from @natfriedman "Imagine we discovered a new continent with 100 BILLION people on it, and they're all willing to work for free!". Read that again. Most people are not realizing this yet. "Yes, I agree @MattPRD, but how exactly do I do it?" Great question. Here's an example. Look at how this ecommerce brand Zox.la is leveraging AI to power a previously IMPOSSIBLE experience: Zox sells bracelets with hidden affirmation messages on them. The problem? They have a HUGE amount of bracelets you can buy, so how do they match you with the right one? This is what they use AI for. On their website they prominently feature an AI quiz (that they made with @OctaneAI) where you answer a series of questions and then get a personalized note and set of product recommendations. Here is an example of one of the questions: The AI quiz collects your email address and saves your quiz answers, so not only does it increase revenue with product recommendations but it increases the brand's marketing list and their targetable data: Check out how my personalized note fully incorporated my answers: And the AI also featured the recommended products from the note where I could one click purchase them: Thousands of people have taken their AI quiz and the note and product recommendations were unique for every single one of them. That is incredible. Not one person saw the same thing. And this is only the beginning of what this technology is going to look like. And it is going to be everywhere. Already thousands of @Shopify stores are using @OctaneAI to power these future forward experiences. I'm going to be talking to the Zox team tomorrow to learn more about how they are thinking about leveraging AI in this new world we are all stepping into. I will share my insights as I find them, and I am so excited to keep improving this technology 🚀

ANNOUNCEMENT: We are going to lock our 90mil @yesnoerror / $YNE dev tokens this week. Stay tuned. Audit science. Accelerate humanity. Seek truth. 👍👎🚫

There may be MANY millionaires made from the upcoming @OpenAI GPT store. You should be one of them. Last night I made a free private community for GPT Builders to share and learn from each other. Already there are HUNDREDS of builders. Want an invite? Reply here. Let's build!

OpenAI plug-ins is the new SEO.

After running many tests, @AnthropicAI's Claude 3 Haiku is better at code than GPT-4-turbo AND its 40x cheaper. You should try building with it. p.s. If you want an easy way to switch between @OpenAI and Anthropic you can use @keywordsai from the newest @ycombinator batch.

🗨️ Ask it to come up with titles For an article, blog post, or book, ask ChatGPT to come up with titles.

What happened with @yesnoerror and $YNE over the past week: - Launched from a 2 hour experiment I posted on X - Over $100mil in volume on day 1 - 10k+ people hold $YNE token - Organically listed on multiple exchanges including @binance Alpha - Telegram community with 1,500 members - Met with @OpenAI team to discuss AI agent system - Public recognition from @gregosuri @Rewkang and more - @pmarca is following the @yesnoerror project - Spent time with our investors at @BoostVC, @BraytonKey and @AdamDraper, mapping out the larger vision - 1,500 research papers reviewed by the AI agent - Flagged 3.5% of papers with errors - AI agent reviews papers 1400% faster pace then when launched - Increased detail of paper analysis - Released plans for the AI Agent Synthetic Data Project - Brought on 4 developers to build the v2 of the AI agent This week: - Building new AI agent capabilities (just wait until you see them) - Increasing scale of papers being reviewed by the AI agent - Putting together a team of A+ advisors - Working on tokenomics and roadmap - Preparing a community of scientists and researches to collaborate on $YNE - Talking to leaders of other DeSci projects - Preparing for a spaces / AMA There is a big vision coming together, what you see right now is less than 1% of what is being worked on. Stay tuned. From everything I am seeing it is clear that DeSci is going to have a very meaningful impact on the world over the next 10 years. Science is about to be completely disrupted and that is good for all of us. This has been a wild wild week ❤️ It is exciting to have the opportunity to play a part in this. Thank you for this astounding support so far. Let's audit all of science. Let's accelerate science. 👍👎🚫

Introducing: BloggyGPT! 1⃣ Send a link to your website 2⃣ BloggyGPT looks at your website, identifies your target audience, your goals, your tone, gathers info on the web, and writes an A+ blog post with you Free for ChatGPT users. Try it yourself here BloggyGPT.com

At @OctaneAI we have made the best no-code product quiz builder for brands using @Shopify (& improving it every day). BUT - Did you know that you can ALSO code quizzes on top of @OctaneAI? Custom CSS, event data and more. Perfect for headless or super premium designs. 👇

While trying to find bugs in research papers for @yesnoerror / $YNE I actually just found a small bug in @OpenAI itself. Just sent it in to their team 🫡

Ex employee at top 2 exchange sent me a 3 page PDF on tokenomics suggestions for $YNE Hedge fund redid our white paper diagrams without us asking Big X account sent me a video they had someone make about @yesnoerror @yesnoerror community is the core of the team. ❤️

FORGET ChatGPT, THIS #AI will replace humans. 🤖 Autonomous Agents Are The Fastest Growing Open-Source Technology In History - But... What are autonomous agents? - How do they work? - How can you use one? I made the best overview on the planet. Your mind will explode 🤯

Good luck @nikitabier 🫡

All intelligence in the world just dropped by 25 IQ points

I started @yesnoerror and $YNE because I saw @pmarca @snewmanpv and @emollick discussing how AI is now good enough to find errors in research papers. So let's go use AI to find all the errors in all research papers? This should totally exist??? So I built it. And it just took off instantly. Mega viral right out the gate. Millions of views, 10k+ token holders, the most viral thing I've ever been involved in. And I dove headfirst into it. What an incredible mission and moment, how could you not? When the world picks you for a mission like this you accept. It's rare that you get to work on something that genuinely can change the world, and this is one of those projects. We've already found errors in over 100 papers, and soon this number is going to go way WAY up. I imagine finding errors isn't the end of this either, I get the sense that it's just the beginning. And it's so interesting, I don't have a background in research or academia, I didn't go to college (I moved to Silicon Valley to join a startup at 19 back in 2007), I just thought this was a really important idea and I knew how to design (and continue to design) and code the AI agentic system that can do this at scale (this I have more experience than 99.99% of people). As a result I have been on a non-stop crash course learning about everything there is to know about researchers, research papers, and the entire process and system behind them. This is all new to me. Turns out a lot of the academic and research system doesn't make any sense at all! And I feel like I'm only beginning to scratch the surface. Incentives are often all wrong, researchers are doing things they don't want to do in order to get funding, journals and other organizations are profiting off of gatekeeping hard work from often times publicly funded research, the list goes on and on. Not only do we need to go find all of the errors in all research ever published, which we are working on, but academia as a whole needs to be completely disrupted. I know how we are tackling the first problem, we've already started, but I feel there may be something we need to do to impact the second one. @yesnoerror may have snapped into existence in an instance, but we're striving for impact that could last forever. I'm saying this out loud, not because we have accomplished this yet, not to make a promise, but because I am manifesting it. This is the goal. If you are involved in academia, if you are at a top university, if you are on the inside of this right now, and you don't think the system is correct, please DM me and tell me the problems we need to fix, tell me what makes you MAD. If you know someone in academia, or who matches the criteria above, tag them here. Spread the word. We're going after this. I don't know who else is? I don't know everything yet, but one thing I promise you is I learn very fast, our team learns very fast, and we know how to build. Let's go.

/Example prompt/ Write a title for a blog post about how to stay productive when working from home.

Imagine AGI/AI is the most powerful intelligence humanity has (and will continue to get better). Imagine it as a being that is floating around the earth and with enough resources you can point it's gaze at whatever you want. Businesses are utilizing it for products and services they can charge for. Which is good and useful for many things. They charge for a service, they make revenue, they use the revenue to build the product and fund the AI. But - What about all of the initiatives that AGI/AI can be pointed at that a business could NOT charge for? What about applications of AGI/AI that are good for humanity but not good for a business? What drives value for humanity but not businesses? @yesnoerror is one experiment to fund this kind of idea. There are millions and millions of research papers in existence. Conclusions in science that humanity has agreed are correct. But humans make mistakes, and we know that AI can catch these mistakes. How do we know AI can catch these mistakes? Because it happened recently. In October 2024 a peer-reviewed study was published stating that black plastic kitchen utensils were toxic and a health risk. People panicked, every publication released articles. People everywhere threw away their utensils. Turns out there was an error in the paper. A number was multiplied by 10x when it wasn't supposed to. And guess what? If you give this paper to o1 from @openai and ask it to find an error, it finds it. In 30 seconds. For $0.30. This experiment is very simple: Use the power of the open market ($YNE) to fund an effort to have AI review past (and future) research papers for errors of many kinds. There will be errors, with the magnitude of papers that exist it is guaranteed. What will they be? What will the implications be? This is an experiment to build an AI that floats around the earth using it's advanced intelligence to check the science of humanity. It's a simple idea, it's a powerful idea, and its a tangible glimpse into our AI future.

Ok I think I follow everyone who works at @openai now

Want to optimize the ads for your @Shopify store? - Run ads to an @OctaneAI quiz - Give personalized product recommendations - Sync all your data to @klaviyo @codyplof is spending $8k/day doing this on TikTok & it's performing VERY well. Learn how 👉 octaneai.com/customers/jones…

Today, we're unveiling Insights Analyst - @OctaneAI's breakthrough AI agent that does research for DTC brands. 1️⃣ Import your reviews 2️⃣ Ask a question or give a task 3️⃣ Get an answer in seconds Analysis, strategy, content, customer voice. This is a total game changer.

What happened with @yesnoerror and $YNE over the past week: - Rapid, deep, and focused thinking on the future of $YNE, there is a much larger vision that I have come to realize as obvious and something that the world needs to have. This is important. - Released v1.0 of our whitepaper - 3,913 research papers reviewed - 1.9% with mathematical errors, 9% with an error of any kind - Over 13.4k token holders - Continuing to build the advisory board. Selected one, evaluating more. We are being *extremely* selective. - Rate of paper review is up. Soon this will skyrocket and what we have done so far will feel insignificant. - High profile interest in @yesnoerror continues to rise as leaders in their fields continue to follow the project, including @pmarca @chamath @SigGravitas @paulkhls @cuysheffield @adaugelli @ericabrescia @Rewkang @gregosuri @BraytonKey @joshelman (and many more which I haven't gone through to find) This week: - Putting the wheels in motion to execute on the larger vision. This requires smart people and speed, both of which we know how to acquire. - Calls with meaningful partnerships. Will share these as they are ready. If you believe we should partner, please reach out. If you *really* believe we should partner, find someone I know and convince them to introduce you to me. - I am leading the dev team on building out v2 of our science paper agentic system, expanding outside of just @arxiv for research papers. - Continuing to run evals with @OpenAI - Increasing the # of research papers reviewed each day - Putting together a larger community of researchers and scientists, and those looking to impact $YNE in a meaningful capacity. - Organizing teams to increase communication to the public of @yesnoerror progress, results, and initiatives. We are auditing all of science. We are accelerating humanity. We are looking for truth. 👍👎🚫

New announcement coming early next week for @yesnoerror & $YNE. Accelerate.

I built @yesnoerror in an evening 2 weeks ago. Now 14k people hold $YNE token. We are using our AI agent framework, powered by o1 from @OpenAI, to find errors in real research papers. This week we announced partnerships with @ResearchHub and @opencampus_xyz What's next?

New @yesnoerror $YNE announcement in next 24 hours.

Growing up I went to a school called Waldorf where I didn't use a computer or watch TV. Right before I went to high school I got a laptop and I became obsessed. 2002: Went to an amazing high school on scholarship. Was surrounded by people who had 100000x more wealth than me, was very strange to go their houses. Fascinated by what their parents did for work to get them into that type of position. 2005: Kicked out of high school because I spent more time building products that doing home work 2005: Hacked the school over summer, but helped them fix it, got back in and graduated. 2007: Was in the early beta of Hulu. I embedded every single Hulu video on my own website and launched it, taking Hulu out of beta overnight. 2007: Produced a live broadcast of someone playing Halo 3 for 72 hours straight, one of the first video game marathons every streamed. Used a small website called Ustream to live broadcast it and took the entire website down after hitting the front page of @kevinrose's Digg. 2008: Met the founders of Ustream after breaking their website, and moved to Silicon Valley to join their team. This was a lucky break as nobody in Silicon Valley was hiring people who didn't have a college degree. Started as an intern doing literally whatever they needed, I didn't care, worked 24/7/365, then moved into business development, and finally into the position I wanted where I was leading product. Got so (so) lucky to be mentored in product by @joshelman, founding team of @linkedin and major impact at @facebook @twitter, and many other platforms - this honestly changed my life trajectory in a way I can't ever really repay, even sometimes by him just telling me that I could do this. Ustream went on to be acquired for 9-figures years later by @IBM (I worked there for almost 4 years). I was young (I joined when I was 19) I didn't really make any money from it, although people probably thought I did. 2011: Bought bitcoin. Lost it all when Mt Gox was hacked. 2012: Took a company through @ycombinator. Was in the same batch as @brian_armstrong when he was the only person who worked at @coinbase. Got advice weekly from @garrytan @sama and others. I built products I was proud of, but ultimately failed to create a business. Felt horrible to fail. 2012: Asked to give the commencement speech at the high school I had been kicked out of. Did the speech. 2014: Built the first crypto social network where you could send bitcoin to each other instead of liking posts, used by every major person in crypto at the time. @BoostVC invested. Notably @VitalikButerin used the platform to do a pre-launch AMA before the ethereum token sale. The idea was a bit too early and was unable to sustain itself. Again, felt horrible. 2016: Started @OctaneAI, an AI chatbots company backed by @BoostVC @generalcatalyst and more (it has evolved quite a lot since then). When I first had the idea for it I called @BraytonKey and @AdamDraper (who by the way @joshelman introduced me to years earlier) and they wrote the first $250k investment check on the spot based on a drawing I had made on a napkin, even though my last company they invested in shut down. They have been with me ever since and I am immensely grateful for the level of trust, support, and dedication they bring. 2018: Focused @OctaneAI on the @Shopify ecosystem, went on to help drive $1bil in revenue for our customers. Actually started to figure things out with my cofounder @benparr, and our team, and built a real business (we are still running it today, profitable and growing). Learned and still learning from @tobi's vision and execution of the Shopify platform. 2020: Was one of the first people invited to be in @OpenAI's private beta before LLMs could even write poetry correctly. Spent every day after that coding with LLMs, exploring different RAG techniques, building experiments of all kinds. April of 2023: I wrote maybe the most widely read article on autonomous agents (over 250k readers). As part of the article I talked to experts like @shl @ericabrescia @tunguz @yoheinakajima @rrhoover @Jihoz_Axie @realSharonZhou @bentossell @MattHartman @jowyang @dunkhippo33 and many more. November 2023: Attended @OpenAI's first ever private dev day, was able to see @gdb @DrJimFan and many others. Was one of the first with access to the ChatGPT app store. Launched multiple GPTs, one of which is still top ranked. 2024: Spent every single day coding with AI using @cursor_ai @ChatGPTapp @perplexity_ai @v0 @Replit. I would sit down and speedrun building apps from zero to something every 30 minutes for hours every night. December 2024: Noticed an emerging trend at the intersection of AI and crypto. Thought it would make for a great article, like my AI agents article. Needed something to build that was an actually useful AI agent that would truly benefit people so that I could talk about it in my article. Saw @pmarca and @emollick talk about the need to use AI to audit science papers and find errors. Used all of my knowledge to build a prototype in a day. Made a token and posted it on X. Instant chaos. Suddenly @yesnoerror + $YNE was born, with a community of over 10,000 people that gathered with a ferocity I have never experienced. Literally out of nowhere in a single day. What was intended as an experiment took on a life of its own. I could either have backed away, or dove headfirst in. I chose the latter. There is something big here. And it is clear I'm not the only one who thinks so. 2 weeks later, here we are. We just put out our first whitepaper. We're building an AI agent to audit all of science. It's already audited thousands of papers. It's already found errors. My timeline isn't perfect. I've failed a lot, and I've learned a lot, but still I am lucky to be put in positions to BUILD, and so grateful for it. Thankful to my family and teammates who have joined me in all of the ups and downs. If I'm in a position to give any advice, then my advice is to go build as well and dive in headfirst. The timing is feeling very right for this concept, some of this is on purpose, and some of this is luck. This quote from @naval has been my north star "Become the best at what you do. Refine what you do until this is true. Opportunity will seek you out. Luck becomes your destiny.". I'm fascinated by the idea that a community is powering an AI product that the world gets the benefits of for free. FASCINATED. There is something so interesting in the underlying fundamentals of how @yesnoerror works. It may cost $30mil to simply power the AI to audit all of the research papers in the world, and no business would ever be created to do this so it would never happen, but a crypto market is actually making it happen. I just can't stop thinking about it. If you can't stop thinking about this concept either, DM me. Diving into this is a risk. Diving into anything is a risk. But to move forward we need to take big risks, and that is exactly what we're doing. What will happen in 2025?

The people are not in control of science. It’s large corps & universities. The public does not have a way to verify any of it. This needs to change. We need to decentralize science and give it to the world. This is a huge movement and @yesnoerror $YNE is leading the charge.

🧪 @yesnoerror can be a defining technology of our new AI world. It started as an experiment, but it is becoming a movement. Here is what the future holds for $YNE: 1. AUDIT SCIENCE There are over 90mil research papers, and millions of new ones published every year. Having a super intelligent AI agent constantly look over our shoulder, checking our work, will help us to produce better science and avoid the fallout from both minor and critical errors made by humans. Inspired by @pmarca, @emollick, and @snewmanpv, our initial focus is to build a system that publicly goes through these for the good of world. We have already started this and actively working night and day to improve and scale it. 2. ACCELERATE SCIENCE Not only can we AUDIT research, but @yesnoerror can play a major role in CREATING research. There are many ways this can happen. Many different companies are creating better and better AI models (@openai @google @anthropic, etc), but most people don't need to use the smartest models available (like o1) because most people are using AI for writing blog posts, writing code for web apps, talking to ChatGPT about an upcoming trip, etc. Even though we have models like o1 that are PHD level models, most people aren't using them! What does that mean? It means that most people building AI products are building for the largest market, which means they are not building technologies with these smartest models. Most people don't need a PHD level AI and don't know what to use it for. So what do we need to do? We need to create AI agents that are PHD level, using models like o1 and above, that are for the smartest people in the world, the scientists. We have an opportunity to leverage all of the learnings from going through these 90mil+ research papers and to create an AI that helps scientists and researchers create NEW science and research that avoids the errors from the past. Every scientist and researcher in the world needs access to a technology like this and we can make it. 3. MAKE NEW SCIENCE AI is getting smarter and smarter. 5 years ago we didn't even have access to a technology like this. Now? AI is everywhere doing everything. But it still has a lot of room to grow on how intelligent it can and will become. How do we know that AI has room to grow in intelligence? Because right now AI just helps us do what we are already doing. It reflects ourselves back. It is smart, and knows what all PHDs in every vertical know, but it is not yet super intelligent. We know that AI has room to grow because it has yet to create new things that people have not already created. We know that AI has room to grow because it has not yet made breakthrough discoveries in physics, or chemistry, or biology. But here's the thing. It will. AI will have more breakthroughs than all of the world's Einsteins combined. Time and time again this is something that @sama the founder of @OpenAI has discussed. @yesnoerror can play a part in reaching this goal. 4. LIVE FOREVER @bryan_johnson is trying to figure out how to NOT DIE. He believes humans can live forever and he has spent millions of dollars reversing the aging process on himself. And he's not the only one working on the science to figure this out, @brian_armstrong the founder of @coinbase has heavily funded @newlimit a company working every day to extend human life. This is a very real initiative and the more we can accelerate science the faster we can achieve this goal. 5. MULTIPLANETARY SPECIES @elonmusk with @SpaceX and @JeffBezos with @blueorigin are trying to get humanity off of one planet and distributed to many, with Mars as the initial focus. This expands our resources, but most importantly, the more planets humans are on the more difficult it is for us to go extinct like the dinosaurs from a planet wide natural (or self inflicted) disaster. It is one of the most important initiatives we have. And we won't stop at being on Earth and Mars, we will then need to continue to expand. The universe is a very large place and if you zoom out we are only at the extreme beginning. The more we accelerate science, the faster we can become multiplanetary. We need to audit science. We need to ACCELERATE science. This is what @yesnoerror is working on.

Are you excited to see @OpenAI’s 30 second Super Bowl commercial this Sunday?! Written entirely by ChatGPT.

Brands using @Shopify have made over $500,000,000 with @OctaneAI quizzes. And now they are getting EVEN more formidable. Introducing A/B testing for @OctaneAI quizzes 🤖📈 Brands using this have doubled their revenue from quizzes. What are you waiting for?

Ok @elonmusk we're integrating @yesnoerror and $YNE with @xAI. Hit me up, let's audit the entire government together with AI. We have the system built to do it. @yesnoerror 🤝 @doge

Had another call with the @ResearchHub president and data scientists yesterday, lots of good stuff happening there ❤️ I’m most excited about the new version of the @yesnoerror framework we’re almost done building. At the end of the day we’re basically building an autonomous AI god / entity powered by $YNE that watches over humanity’s scientists and researchers, and the tech is just so exciting. Both from how it works but also the impact it’s designed to have. And I say that from the perspective that I was one of the first people in the world to be building with LLMs over 4 years ago, and almost 20 years experience building in tech in Silicon Valley. We are pointed in a meaningful direction. I know we may all have gotten here for different reasons, but no matter the reason we did somehow get gifted this unique mission and we’re all going to see it through. I appreciate all of you. I do not know everything, but together we might know most things. Whatever you can do to learn about this space, this mission, to have ideas, to contribute in any way you are able, I encourage all of us to do so. We are together and unified. Onwards. 👍👎🚫

The new @yesnoerror AI upgrades are insane. 1,000x improvement. This is a super agent we can point to find errors and problems in ANYTHING. @RobertKennedyJr, you want to audit every mRNA paper? Done. @bryan_johnson, you want to audit every longevity research paper? Done. @elonmusk, you want to audit every government document? Done. 👍👎🚫 $YNE

New $YNE / @yesnoerror info being released in the morning PST.

We launched $YNE and @yesnoerror on @pumpdotfun and @X 43 days ago. Out of the 6.9 million tokens ever launched on @pumpdotfun, $YNE is ranked #25 by highest market cap, and we continue to hold strong in that range. Turns out it's a bit abnormal to launch a serious project this way. We did it as an experiment. The experiment got real big real fast. We are one of the only real products in the space, surrounded by memes. My original goal was "what if we built something that was actually good for the world and tokenized it?". That ended up being a huge differentiator. What makes @yesnoerror so different in the Solana and crypto ecosystem? A few things: - PUBLIC GOOD: We are building an AI agent that offers a public good. We're using AI to audit scientific research, first to find errors, but more on the way. - REAL TECH: We have been building with AI since 2016, we were one of the first few hundred teams to have access to @openai's first beta in 2020. We have been building with LLMs, RAG, and autonomous agents since it was possible to do so. - DOXXED + REAL CONNECTIONS: In a world of crypto, cabals, anonymous actors, vaporware, and memes, myself and my cofounder @benparr are known people who have been building real businesses, and real relationships and experience in Silicon Valley and tech for almost 20 years each. When @yesnoerror blew up I called @BraytonKey from @BoostVC and he fired me up and said this project was incredible. How amazing to have people like that who prop us up and propel us forward. - REAL BUSINESS OPPORTUNITY: This started as a public good, but we have also discovered that disrupting academia and science is big business. Turns out a lot is broken here, and @yesnoerror is in a fantastic position to help fix it and accelerate innovation. We can build a real business and generate revenue, not from selling tokens, but from offering subscription products to students and researchers to help them make better more accurate research papers. And this is only the start, there are many other interesting ways to monetize what we are building. Public good on one side, real business on the other. Having 14k+ holders is a very unique and interesting way to speedrun building a product. 14,000 people asking every day "what did you get done today?" 14,000 people who are here to market the product 14,000 people making introductions I don't think building in this type of environment is for everyone, but I see so much potential and good in it. I love it. I think for the right team, tokenizing your product can be a superpower. We're 43 days in. Where do you think we will be at day 60?

I built myself a little realtime crypto AI alpha dashboard powered by DeepSeek + X.

Gone are the days of solitary shopping. With @iliabeauty + @OctaneAI, welcome to the era of collaborative beauty decisions.

The @yesnoerror team's 90 million $YNE tokens have been locked until May 1st, 2025. Let's audit all of science with AI. Onwards.

🗨️ Get unique examples Ask ChatGPT for examples to illustrate a point or concept.

Great to see @yesnoerror and $YNE in the official DeSci report from @binance

$YNE Update: We will be implementing a cheaper "Suspicion Score" on @yesnoerror where we can use low cost LLMs to rank suspiciousness in each research paper, and then multiply that by citation counts and fields ranked by how much they impact the general population. This will allow us to prioritize the papers we go through to maximize discovering errors that have wider impact.

📊 @yesnoerror / $YNE is officially listed in the DeSci and AI Agent categories on @coingecko

🧪 Introducing the @yesnoerror $YNE AI Agent Synthetic Data Project. This is the system we have put together to teach our AI agent how to be better and better at identifying errors in research papers. It is a very powerful system that allows for the AI agent to get smarter and smarter the data we introduce it to. 1. We collect a large number of existing research papers from various topics. 2. We use o1 from @openai to make changes to each paper to introduce minor or critical errors. 3. The @yesnoerror AI agent is then set to go through each paper, not knowing ahead of time what the errors are, with the goal of determining if there is an error and if so what it is. 4. Because we already know what each error in each paper is we look at what error @yesnoerror found, if any, and we use AI to check if it is the same error that we know to exist (because we added the error ourselves). 5. After going through all of the synthetic errored papers, we are able to give the AI agent a score. For example did it detect 82/100 errors correctly. 6. We then update the AI agent's knowledge, and thought process, based on the errors it missed, to better equip it to catch those errors in the future. 6. The process is then started again and a new score is achieved. This process is repeated until the AI agent is able to get 100/100 every time. This is some insight into how we are building out an AI to go through all 90mil+ research papers ever published.

This is an autonomous agent ordering PIZZA in the browser. Just tell it to do whatever you want... and the AI figures out how to do it and DOES IT. This is the future and we're really only in the very very beginnings of what this is going to look like. #AutoGPT #BabyAGI

Had a great brainstorm about $YNE & @yesnoerror with @gregosuri founder of @akashnet today. He has built a massive community around $AKT, one of the largest AI blockchain projects. Listed on @coinbase @kraken and all the other top exchanges. Thank you for your advice today 🤖

Apple's Vision Pro headset is a replacement for your computer and monitor, not a replacement for your Xbox or Playstation. Going from 1 or 2 physical monitors while working to completely immersed + ability to tune out outside distractions = huge productivity gains.

Tbh this might be how everything works

We added @ResearchHub as an official @yesnoerror partner + @benparr is in Washington DC talking to Congress and other leaders about AI, accelerating humanity, and $YNE. This is nothing. This is just groundwork. Acceleration is approaching.

Over 5,000,000 coins have been launched on @pumpdotfun. After 11 days @yesnoerror / $YNE is in the top 50 of all time. I haven't slept and have been working non stop to become an expert at the ecosystem, everything from @ai16zdao to @virtuals_io to @pumpdotscience @truth_terminal @0xzerebro and more. No sleep, but full of energy and hooked directly up to the firehose here. There are a lot exciting innovations happening in this ecosystem that I don't think most people in the world know about. Being on the fringe is a weird but fascinating place to be. pump.fun/coin/7D1iYWfhw2cr9y…

Introducing: Simpsonize Me GPT! 1⃣ Upload your profile photo 2⃣ Simpsonize Me GPT turns you into a Simpsons character I made this with @OpenAI's new ChatGPT creator in under 10 minutes. Link to try it yourself below 👇 Let me see your Simpsons pictures!