Mike Remondi

You know your startup is doing well when it's Saturday at 8am and you wake up to an alert that your DB is pegged at 100% CPU. Just @ethanhlo and me doing some fun DB surgery to start the weekend off strong 🫡

@pmarca “How can I not build when it’s this easy to build and there’s so much to build?”

True, and the exact opposite of the constant claims to the contrary, but precisely what you’d expect from a huge surge in productivity.

Thank god… @salesforce going headless. Beginning of the end for one of the most excruciating user experiences ever created

@zoink Figma make has been unimpressive and really has left the door open to a ton of competition. Both on the prototyping side (replit, etc) and on the design side (cursor, pencil, etc). Really feels like an opportunity that Anthropic is well positioned to take advantage of

@ClaudeDevs I really like how @AmpCode gives you the dollar cost of a session in line in amp. That would be a super useful, more real time feedback loop of $$$ usage too

We’re adding more visibility into where your Claude Code usage goes. Run /usage to see a breakdown of what's driving it: parallel sessions, subagents, cache misses, long context, plus tips to optimize each.

AI attacks are cheaper to run, infinitely scalable, and personalized at a level that makes gut-feel detection useless. @AdaptiveSec protects your org from these new AI threats. Shoot me a DM and we'll build a deepfake phishing simulation of someone at your company!

@samlambert Coding is dead; long live software engineering x.com/remondimi/stat…

OH: AI agents are great at writing code but they're terrible at engineering.

@rezoundous More mythos marketing 🙄

Why does Opus 4.7 keep checking for malware.

@businessbarista @da_fant YES -- if I were starting my own company right now, I would probably mandate github for all of our knowledge base with regular syncs of external data (emails, etc) to provide rich context. Only issue is that git has no notion of RBAC or permissioning built into the filesystem

Someone is going to build a worldclass “Brain” for enterprises & make a stupid amount of money. Why? As @da_fant said, “coding w ai is solved bc all context is in the git repo. knowledge work is difficult bc context is spread out. an ai system that creates a git repo w all context for a knowledge worker will be able to 100% automate the work.” When companies talk about being data ready for AI, this is what they’re implicitly saying. Engineering has been prepared for this moment for a long time because of the deterministic nature of code, the centralization/versioning of data (read: GitHub), and AI tools that are largely build by engineers for engineers. But for the rest of white collar work, there’s a TON of catching up to do to properly harness the power of the technology. The big challenge here, and why no one has truly cracked the code for "an ai system that creates a git repo w all context for a knowledge worker" is because unlike code, most knowledge is 1) distributed, 2) unstructured, and 3) unverifiable. It's distributed: transcripts live in Granola. Documents in Notion. Customer Data in Hubspot. ERP. Emails. Slack messages. Random spreadsheets. SOP docs. Etc. Etc. Building an ingestion engine that connects to all of your disparate data sources and auto-updates based on the shelf-life of the data is the first, and frankly, easiest step of the process. Next, it's unstructured: let's say I want to create a proposal for a potential client. To nail the proposal, I want it to pull important information from a variety of sources. The specific asks & background from our initial sales call. Previous proposals to anchor ourselves to a proven format. And completed sprint boards from Linear, so the pricing & timeline in the document is grounded in truth. Whether it's a thoughtful filesystem (a la Obsidian) or an OpenClaw-esque memory structure, the brain needs to be great at self-organizing in a thoughtful schema. This is very hard, especially if you want to build a generalizable brain that can be shaped to an array of different enterprises. And finally, most knowledge is unverifiable: writing a function, running a unit test, and seeing if the code works is easy. It works or it doesn't. Using AI to accelerate your content creation process is highly subjective. What is a good/bad idea? Is the content in your voice or not? Does it feel like slop or novel? Answering these questions are both difficult and non-verifiable. That same system described above doesn't just have to be great at organizing & forming coherent relationships, but it also has to be great at self-improving based on feedback from the user. Memory systems (like those introduced by OpenClaw) are great to a point, but as you scale the corpus of data within your company's brain, things like compaction and cleaning become wildly important to avoid the needle in the haystack problem. Someone is going to figure out how to solve this problem, and when they do, not only will they make a shit ton of money, but they'll be robinhood for knowledge workers, enabling non-engineers to enjoy the sort of leverage that only technical folks have felt for the last few years.

@jxnlco This is the single best thing I've done for my productivity. Takes a long time to set up properly but if you commit to using it, it becomes SO useful over time. Honestly has become a great thought partner for strategy docs since I dump so much context into the repo

5 steps to making codex your chief of staff 1. download the desktop app 2. install the plugins you need for work 3. paste this into a thread and pin it 4. ??? 5. monitor the situation gist.github.com/jxnl/e96b08aae…

@gdb The arms race between @claudeai / @bcherny and @OpenAI / @gdb is going to result in some amazing products. Very excited to give Codex another try

Codex is becoming a turbocharged partner for everything you want your computer to do for you:

We know better than anyone: we've built an AI and OSINT first phishing simulation product. Sending highly personalized, compelling phishing attacks at scale is trivial and the existing tools have simply not kept up

Phishing is up 1,265% and half of it gets past Microsoft and Google filters. Legacy email security was built to catch known threats: bad links, bad domains, known payloads. AI-generated phishing doesn't look like any of that. We built Adaptive Email Security from scratch to catch an entire new generation of threats

@Cloudflare Been waiting for this. Search is the missing piece for most agent workflows. We stick with file directory structures for LLMs - way better than vector search. But having search as a primitive for other data types changes everything. Curious about performance vs rolling your own.

AI Search is the search primitive for your agents. Just create a search instance, upload, and search. cfl.re/4chlcwQ

@gdb We're rolling out a "company operating system" with Agents & context bundled together across all of @AdaptiveSec. This is the typical "aha" moment that we try to get people towards on day 1 of using it

always a real feeling of magic to ask codex to perform a task that requires finding information scattered across slack, google docs, notion, and various internal tools, and it just figures it out

tired: we ship so fast at the cost of uptime expired: we have five 9s of uptime wired: 🌶️🫑🫑🌶️🌶️

We built our own Agent Builder about a year ago on top of @temporalio and deeply integrating it with our own tool calls and data. This is interesting... but the ability to switch to any model provider and build our own functionality on top of it still seems too valuable to switch to something like this

Build long-running agents with more control over agent execution. New capabilities in the Agents SDK: • Run agents in controlled sandboxes • Inspect and customize the open-source harness • Control when memories are created and where they’re stored

AI is being weaponized against our least prepared family and friends. We at @AdaptiveSec are doing as much as we possibly can to educate the elderly on what LLMs are capable of and how to stay safe. It's as important a mission as I've worked on in my career and I'm incredibly proud of the impact we're making

@Polymarket Slop is what happens in the period between trying a new tool and mastering it. Part of the growing pains and the meme of "workslop" just slows down the progress

JUST IN: Use of AI in the office is reportedly creating a flood of “workslop” that takes longer to fix than do from scratch.