Greg Kamradt
@GregKamradt • 47,262 subscribers
President @arcprize, builder/engineer
Shorts
Videos
The mindset shift that changed the way I looked at the world: “I’m going to build my own tools” I've automated more daily workflow in the past 12 months than I have my entire career “You can build your own things that other people can use. Once you learn that, you'll never be the same again.” - Steve Jobs "Coding is dead" - Said no one serious The secret wasn’t running away from code No! The opposite. I dove head first into new tools (Cursor, v0, LangSmith, OpenAI, Anthropic, etc.) As the CEO of Cursor puts it, we are entering the golden age of “delight concentration” We can now be creative instead of technical. The friction to building your own tools is literally just a prompt away (Anthropic Artifacts). For custom tools, the barrier to entry is mind-boggling low. Tools like Cursor, v0, shadcn, LangChain and Vercel allow you to build, deploy and share what matters to YOU faster than ever It’s no longer skill level that's holding you back, it’s just your desire to take a step forward
Greg Kamradt117,317 次观看 • 1 年前
Well, it happened. I accidentally became obsessed I found a business that built a Q&A app using Slack messages as a knowledge base Years of tribal knowledge locked behind keyword search LLMs make it useful Way bigger opportunity than I thought Here's 40 companies building:
Greg Kamradt81,389 次观看 • 2 年前
Low key one of my favorite presentations of the year Sam Whitmore sharing her "AI Memory" journey building New Computer * How her AI Memory/RAG approach changed 2023 > 2025 * Why the perfect memory architecture doesn't exists * 4 types of memory (queried in parallel)
Greg Kamradt33,972 次观看 • 11 个月前
How good is GPT-4-Vision at extracting text from images? I wanted to find the limit - but I found weirdness instead Most surprising: GPT-4V performance varies depending on the *structure* of text it sees Let me explain A set of images with progressively more text was presented to GPT-4-Vision. GPT-4V was asked what text it saw in the image. The response from the model was compared against the image’s original text and scored for similarity. The model was tested with 4 types of text: essay, random words, random tokens, and random characters. Findings: * Performance degrades - Yes, the models are good at basic OCR, but as you get more text and words then performance drops (this is expected) * Type of context matters - You should expect different recall on your texts based on your context types * Hallucination Errors - I thought that the model would make errors of omission (it wouldn’t return all the words). But instead the model mostly made hallucination errors - it replaced words with made up words. * Evals Matter - This test in isolation doesn’t mean that your data will have the same results, but it should motivate you to create eval tests for your data and anticipate errors which are hard to spot Notes: * Next step would be to add additional image types like tables or PDFs * GPT-4V would routinely get stuck in repeat-token-loops when trying to extract random tokens * GPT-4V would refuse to answer most random character images
Greg Kamradt49,109 次观看 • 2 年前
Just tried ChatGPT agent on ARC Prize ARC-AGI-3 > Told it to play a game > Couldn't figure out what to do > Agent did a web search, "how to beat X arc-agi-3 game" > It didn't find answers > I told it to try clicking red/blue blocks > It clicked them, noticed something happened, kept clicking > Nudged more > Couldn't figure it out > Searched again Then I cut it off btw agent is a very cool tool
Greg Kamradt15,380 次观看 • 10 个月前
没有更多内容可加载