If you want to understand #AI beyond the hype, this is the room to be in!

Judging by my tl there is a growing gap in understanding of AI capability. The first issue I think is around recency and tier of use. I think a lot of people tried the free tier of ChatGPT somewhere last year and allowed it to inform their views on AI a little too much. This is a group of reactions laughing at various quirks of the models, hallucinations, etc. Yes I also saw the viral videos of OpenAI's Advanced Voice mode fumbling simple queries like "should I drive or walk to the carwash". The thing is that these free and old/deprecated models don't reflect the capability in the latest round of state of the art agentic models of this year, especially OpenAI Codex and Claude Code. But that brings me to the second issue. Even if people paid $200/month to use the state of the art models, a lot of the capabilities are relatively "peaky" in highly technical areas. Typical queries around search, writing, advice, etc. are *not* the domain that has made the most noticeable and dramatic strides in capability. Partly, this is due to the technical details of reinforcement learning and its use of verifiable rewards. But partly, it's also because these use cases are not sufficiently prioritized by the companies in their hillclimbing because they don't lead to as much $$$ value. The goldmines are elsewhere, and the focus comes along. So that brings me to the second group of people, who *both* 1) pay for and use the state of the art frontier agentic models (OpenAI Codex / Claude Code) and 2) do so professionally in technical domains like programming, math and research. This group of people is subject to the highest amount of "AI Psychosis" because the recent improvements in these domains as of this year have been nothing short of staggering. When you hand a computer terminal to one of these models, you can now watch them melt programming problems that you'd normally expect to take days/weeks of work. It's this second group of people that assigns a much greater gravity to the capabilities, their slope, and various cyber-related repercussions. TLDR the people in these two groups are speaking past each other. It really is simultaneously the case that OpenAI's free and I think slightly orphaned (?) "Advanced Voice Mode" will fumble the dumbest questions in your Instagram's reels and *at the same time*, OpenAI's highest-tier and paid Codex model will go off for 1 hour to coherently restructure an entire code base, or find and exploit vulnerabilities in computer systems. This part really works and has made dramatic strides because 2 properties: 1) these domains offer explicit reward functions that are verifiable meaning they are easily amenable to reinforcement learning training (e.g. unit tests passed yes or no, in contrast to writing, which is much harder to explicitly judge), but also 2) they are a lot more valuable in b2b settings, meaning that the biggest fraction of the team is focused on improving them. So here we are.

This will likely lead us to #AGI !

I got access to ManusAI and it’s mind blowing good. I do not get paid for promotion. I just tried preview access and it outperformed OpenAI’s DeepReseach by a lot! Mind blowing x.com/jenzhuscott/status/189…

Has anyone benchmarked how much these AI tools like Perplexity, Grok or ChatGPT with reasoning mode (and also web search) would score in a JEE Advanced? If someone’s interested in carrying out a rigorous evaluation, would be happy to support with API credits.

The Model Context Protocol (MCP) is not just "another API lookalike." If you think, "Bro, these two ideas are the same," it means you still don't get it. Let's start with a traditional API: An API exposes its functionality using a set of fixed and predefined endpoints. For example, /products, /orders, /invoices. If you want to add new capabilities to an API, you must create a new endpoint or modify an existing one. Any client that requires this new capability will also need modifications to accommodate the changes. That issue alone is a colossal nightmare, but there's more. Let's say you need to change the number of parameters required for one endpoint. You can't make this change without breaking every client that uses your API! This problem brought us "versioning" in APIs, and anyone who's built one knows how painful this is to maintain. Documentation is another issue. If you are building a client to consume an API, you need to find its documentation, which is separate from the API itself (and sometimes nonexistent.) MCP works very differently: First, an MCP server will expose its capabilities as "tools" with semantic descriptions. This is important! Every tool is self-describing and includes information about what the tool does, the meaning of each parameter, expected outputs, and constraints and limitations. You don't need separate documentation because the interface itself is that documentation! One of my favorite parts is when you need to make changes: Let's say you change the number of parameters required by one of the tools in your server. Contrary to the API world, with MCP, you won't break any clients using your server. They will adapt dynamically to the changes! If you add a new tool, you don't need to modify the clients either. They will discover the tool automatically and start using it when appropriate! But this is just the beginning of the fun: You can set your tools so they are available based on context. For example, an MCP server can expose a tool to send messages only to those clients who have logged in first. There's a ton more, but I don't think I need to keep beating this dead horse. AI MCP > AI API *micdrop*

Excellent list of #AI tools every developer must learn

New Alexa! Powered by an LLM and actually able to function as an agent to do things for you

This is my Figma to SwiftUI code workflow I couldn't find a good way to turn my Figma designs to Xcode so I built a plugin. It took roughly a day to create with Claude AI. A few more days to add additional features like Save file, Fast forward, Preview options, advanced settings, a ton of testing and of course shipping/feedback. We're at a point where we can built our own tools to complete a workflow.

We’ve also updated our license to allow developers to use the outputs from Llama models — including 405B — to improve other models for the first time. We’re excited about how this will enable new advancements in the field through synthetic data generation and model distillation workflows, capabilities that have never been achieved at this scale in open source.

That's a huge boost for open-source #opensource #ai

Starting today, open source is leading the way. Introducing Llama 3.1: Our most capable models yet. Today we’re releasing a collection of new Llama 3.1 models including our long awaited 405B. These models deliver improved reasoning capabilities, a larger 128K token context window and improved support for 8 languages among other improvements. Llama 3.1 405B rivals leading closed source models on state-of-the-art capabilities across a range of tasks in general knowledge, steerability, math, tool use and multilingual translation. The models are available to download now directly from Meta or @huggingface. With today’s release the ecosystem is also ready to go with 25 partners rolling out our latest models — including @awscloud, @nvidia, @databricks, @groqinc, @dell, @azure and @googlecloud ready on day one. More details in the full announcement ➡️ go.fb.me/tpuhb6 Download Llama 3.1 models ➡️ go.fb.me/vq04tr With these releases we’re setting the stage for unprecedented new opportunities and we can’t wait to see the innovation our newest models will unlock across all levels of the AI community.

Check out my latest article: Anthropic's (Claude) Artifacts: A shift in the AI Leadership? linkedin.com/pulse/anthropic… via @LinkedIn #AI #TechInnovation #Anthropic #ClaudeAI #ArtifactsFeature #ArtificialIntelligence #OpenAI #ChatGPT #RapidPrototyping #FutureOfTech

Great resources for anyone wanting to start learning CUDA, thanks to @mallocmyheart for sharing

⚡ Can't believe this worked so well. From a SketchUp basic render to a super high resolution "AI render" in 7 minutes. Using the 3 Magnific spells together: 🎭 Style Transfer 🔆 Relight 🪄 Upscale These are the steps I followed so you can do it too 🧵👇

2/ DispatchAI AI agent 911 dispatchers that immediately respond to crisis calls, measure caller anxiety/sentiment with @hume_ai, and send emergency responders on site AI agents for crisis response 🎖️ AI for Good Award 🥇 @intel Prize 1st place 🏆 Grand prize winner