Antonio Silveira

@jasonlk Have you played with the new version of Codex App? It is not focused on design, but the new feature of commenting on the web app and ask for changes is really cool.

Ok my quick verdict on Claude Design: - It's better than what you get out of Opus 4.7 - It seems pretty powerful already - It has an asset store etc which is great But - It still looks like "Claude" - It's not very creative out of the box Not sure I'd start here for a project, but others might. I'm always tight on time and impatient so for me I may not use it. Not yet. For a website, not sure this is a game changer ... yet.

@jasonlk Great stuff Jason, Subscribed on YouTube.

The Agents Ep #001 is Out!! Three things that broke in our agents this week. None of them were our fault. All of them were our problem. 1⃣Preview environment outage. The AI Agents blamed each other. Multiple apps lost database connectivity. The agent tried to debug it and blamed Qualified (not the issue). Then blamed other integrations (also not the issue). Production was fine, but we couldn't iterate for hours. 2⃣Micro hallucinations in our AI VP of Marketing Yesterday it said we were 44% ahead of plan. This morning it said 11%. Same agent, same data. When I asked what happened, it said: "Oh, I was comparing to the wrong year. And since I didn't have the right year, I made up the data." 3⃣Model regression in our AI VC pitch deck analyzer. Over 4,000 decks graded. Stable for months. Then without anyone changing a line of code, it started telling every startup they had $100K in revenue growing 500%. A silent Claude model update broke a complex multi-step workflow. We now spend 15 minutes a day maintaining each of our AI agents. Without that daily maintenance, agents drift. Slowly. Quietly. Further from reality. Getting your vibe coded app into production is like closing a sale. It's the start of the journey, not the end. Who on your team is maintaining your agents? From Episode #001 of The Agents

For a long time, we have been talking about the models. I have been reading a lot about the lessons from the Claude Code leak and breaking them down into a few areas. The first one is the agent harness. One of the most interesting lessons from the source code leak, at least in my opinion, is that the real product is the harness around the model. The model call itself is one part of the agentic system. If you already have an LLM provider that satisfies you, what matters in practice is everything else around it: prompt assembly, memory, tool routing, retries, permissions, session continuity, and all the operational work required to make agents reliable. I have seen the same thing while building my own agents using @openclaw and other harnesses. Recently, after @AnthropicAI changed the policy around using the Claude Max subscription, I switched my agents to GPT 5.4 through my @OpenAI Pro subscription. The harness stayed the same. The configuration stayed the same. And that is exactly the point: what makes the agents useful is not just the model. It is the surrounding harness 🤖 Here is what I wrote about it in part one 👉 antonio.md/posts/claude-c…

Very good article from @hwchase17 about the agent harness and memory achitecture.

Nice, I hadn’t noticed that @openclaw released a /btw command like @claudeai Code. Very handy when running long, complex sessions.

⭐️ Saturday mornings are perfect for learning new things. 💡 One of the biggest problems in agent engineering is managing context and memory, especially in long, complex sessions. All agents use some form of compaction. It helps keep the session alive, but it is lossy. Important details get flattened, and execution quality gets worse over time. All the frontier LLM companies are working on this, but most of it is happening behind closed doors. That is why it is interesting to watch what 🦞 OpenClaw is doing with its new Dreaming feature. I think it is a real step forward. It treats memory as something that needs curation, promotion, and hygiene, not just bigger context windows or gigantic memory files. I wrote more about it here 👉 antonio.md/posts/why-agen…

Not all agent work is equal 🤖🦖🕵️ The key question is not just what an agent can do. It is whether the human operating it can verify the output with confidence. Some agent work is easy to check. Some looks good on the surface but is much harder to judge. In my latest post, I build on Bian's framing and add another lens that matters a lot in practice: impact radius. Worth being explicit about whether you are building with agents. face, but is much harder to judge. In my latest post, I explained this framing and added another lens that matters a lot in practice: the impact radius of a mistake. Worth being explicit about if you are building with agents. I wrote more about it here: antonio.md/posts/not-all-…

My experience so far is that Anthropic tends to be more direct and efficient. OpenAI tends to be better at keeping you moving and never leaving you at a dead end. Both can be useful, but they create very different interaction patterns, and that changes the agent behavior.

Now that @AnthropicAI is blocking using the Claude subscription with 3rd party harnesses, I moved all my Agents to use @OpenAI GPT 5.4. What happens to my agents was very interesting. Is Agent personality a thing? I wrote more about that here 👉 antonio.md/posts/model-be…

@saranormous Haha, my team created stickers with that on it.

@drewwilson Meta Threads?

I think I've got it. X is no longer a social network first. X is all about hearing from randos and what's in the zeitgeist. A virtual town square. Twitter was about hearing from your friends and people you wanted to hear from. A social network. So, where's the next Twitter?

Great new releases from @OpenAI today at the #OpenAIDevDay event. ChatKit, AppsSDK, AgentKit, tons of updates to Codex and New APIs

Just saw one of the new @Waymo cars in SF, finally ditching the Jaguars for a custom self driving design.

@wesbos It printed really well, I had to deburr the corner of the button to remove some of the support lines. But otherwise great model.

@wesbos Thanks man, love it, I will start printing it right now

These make me very happy

@CaltransDist3 @CountyElDorado @PlacerCA @NevadaCountyCA @CHP_Truckee @CHPSouthLake Any ETO for when I-80 will open?

If you're traveling in the Sierra tonight, expect slow sledding. I-80 traffic is being held in both directions due to spinouts and chain controls remain in effect on Highway 50 in @CountyElDorado. @PlacerCA @NevadaCountyCA @CHP_Truckee @CHPSouthLake

I am finally able to test the new @Tesla FSD 13.2.2 and it is really a major leap when compared to where the software was just 12 months ago. It is unbelievable how much more confident and agile this version is. Wow.

🍎 @Apple finds itself in a predicament. I’ve been using Apple Intelligence with iOS 18.1 📱 for quite some time now, and while some features are promising, they lack the compelling power to truly delight users. ✨ Writing Tools is the standout feature, but its capabilities are very limited. AI photo editing 📸 falls short of the sophistication offered by Google Photos. Siri has improved, especially with better contextual understanding, but after using @OpenAI ‘s advanced voice mode, I expect that Siri should be able to have a long, open-ended conversation with me. Instead, when you try to chat about anything beyond basic questions, Siri presents you with a @Google Search result 🔍 and then the conversation stops. That is really bad— I don’t want a search result with links to sites; now I want the answer. This partnership with Google is undoubtedly lucrative 💰 for Apple, but it also hinders their ability to innovate and deliver a truly exceptional user experience. #AI #Apple #OpenAI #Siri

@adityaag Same here, and I am running ios18.1 beta which has the Apple Intelligence features and to say they are underwhelming is a massive understatement. The camera button is also a gimmick.

I "upgraded" from the iPhone 14 Pro to the iPhone 16 Pro. I literally cannot tell the difference. It took me 24 hours to set up the new phone properly etc. It just feels like a waste of time. And I do not understand where this "Apple Intelligence" is????