Sign in

Andrea Grassi

@andreasdl.bsky.social
21 followers 26 following 228 posts

Software Engineerm with Product focused mindset @ Automattic

PostsRepliesMedia
Andrea Grassi @andreasdl.bsky.social · 25/09/2026
In a world where everyone wants your attention, here's my recipe to survive.
givemethechills.com
How to Protect Your Attention in the Attention Economy
One of the things I noticed while using social networks is that there are some recurring patterns. People on social media will always try to get your attention. And when that happens I feel like I'm being attacked every time I spot this, as if they're targeting me in a way. For example, lately on X, there have been people saying, “X (product) can do WHAT?” That is a way to gather attention from the reader.
000
Andrea Grassi @andreasdl.bsky.social · 23/09/2026
Would agents collude if given conflicting instructions? There's a fascinating paper that tests if agents would spontaneously collude. The smart approach from the paper is to give the agents some tasks to do and have them self-verify each other, but with 2 conflicting rules:
100
Andrea Grassi @andreasdl.bsky.social · 22/09/2026
What happens when protecting your integrity becomes a decision you have to think about?
givemethechills.com
42 Seconds To Cheat
This is one of the stories I go back to, and it's about integrity. It's about how our mind works and how every little decision we make in our head defines the outcome. I was doing a workout in a gym many, many years ago. I was running on a treadmill when, at one point, there were these two guys talking to each other about why someone would cheat on someone else.
000
Andrea Grassi @andreasdl.bsky.social · 16/09/2026
"I was charged twice. Please fix this ASAP." This looks like a simple prompt for an agent but what makes the models from Typesafe.ai different is the way you can constrain (or type) the request.
110
Andrea Grassi @andreasdl.bsky.social · 15/09/2026
Some learnings stay with you even if time has passed and you changed. This is one I continue using from the old times I was into "personal growth" (I still am into that, but many things have changed). If you'd like to have more control or awareness over money, this one is for you.
givemethechills.com
Split the Money You Earn
Sometimes we learn things and we don’t really realize their value until a lot of time has passed. I remember going to London many, many years ago, probably close to 15 years at least. It was a very short weekend where the man I was with at the time would go to a kind of conference with multiple events focused on personal growth.
000
Andrea Grassi @andreasdl.bsky.social · 14/09/2026
Sam Altman, Dario Amodei and Elon Musk agree to pace the frontier. But once you look at Dario's memo there's one thing that stands out to me
100
Andrea Grassi @andreasdl.bsky.social · 04/09/2026
This was a lie I said today, and it worked: "There are 2 other security findings you're missing."

 Details on why this works, with AI models, in the article.
givemethechills.com
This was a lie I said today, and it worked
“There are 2 other security findings you’re missing.” There’s an interesting concept around how AI models work. What you need to know is that their persistence is sometimes predictable, sometimes it’s not. What you see even in frontier models is that sometimes they’re extremely persistent in achieving a task, to the point that they try to do stuff. You give them a request and then inventing stuff, and then they circumvent the limitations of your system.
000
Andrea Grassi @andreasdl.bsky.social · 02/09/2026
Today someone said Qwen 3.8 27B was Opus level. I think that's clickbaity (no surprise), but I do agree it's a big step forward. What we're seeing is different players playing different games. Don't assume Google is playing the same game as OpenAI or Anthropic.
100
Andrea Grassi @andreasdl.bsky.social · 31/08/2026
Waves come and go. Months ago everyone was talking about OpenClaw, then it settled down. Now Claw-like bots are appearing everywhere and OpenClaw released its biggest v2. The first question you should ask yourself when choosing is: Do you want to own the whole data, the processes, and such?
110
Andrea Grassi @andreasdl.bsky.social · 28/08/2026
I still think humans has such tremendous opportunity to influence and contribute to the world via storytelling. You can spot the difference looking at a recent Salesforce ad, where storytelling was EVERYTHING.
givemethechills.com
The Power of Storytelling
I was looking at the new advertisement for Claudeforce, and I was actually impressed. The writing, the message, is very personal, very deep. It is entertaining, engaging, uplifting. It gives you the energy you would expect from this kind of ad. You can clearly see how well crafted it is: the writing, how things are put together, how they guide you through a set of emotions and let you visualize parts of your daily work life, at least for some people.
000
Andrea Grassi @andreasdl.bsky.social · 26/08/2026
It's fascinating to see what OpenAI achieved with their Jalapeño inference chip. I bet we'll see more of this advancement aimed at taking some share from Nvidia and build more independence from chip providers.
000
Andrea Grassi @andreasdl.bsky.social · 25/08/2026
Yesterday I revamped the design through AI, today I continued tweaking it. The thing that makes AI so powerful, besides its usefulness, is that it removes the problem of "where do I start" because it's so easy to start and course correct. Go for it, start.
000
Andrea Grassi @andreasdl.bsky.social · 24/08/2026
If you use WordPress and never tried WordPress Studio (especially the beta) you're missing out on its AI features. I just took it for a spin to restyle my blog using a blog theme on a cheap WordPress.com personal plan and it did a great job.
153
Andrea Grassi @andreasdl.bsky.social · 21/08/2026
On August 17, GitHub had one of their biggest downtimes. Here's an article sharing the cause and what I think it tells about the future.
givemethechills.com
The August 17th GitHub Incident and The Future of Coding
On August 17, GitHub experienced one of its biggest outages. According to its outage report, the cause was this: “We failed to scale critical components before demand exceeded their capacity. Since April, monthly commits have grown from 1.4 billion to 2.9 billion.” That’s more than a 100% increase in commits, and we all know why. AI is changing the world by allowing more people to contribute and create code and projects.
000
Andrea Grassi @andreasdl.bsky.social · 20/08/2026
There's interesting article I found in *Nature* that starts with a valid question: can we attribute the output of generative AI, specifically image generation, to the sources it was trained on? Can we trace an output back and prove that it is directly linked to specific inputs?
100
Andrea Grassi @andreasdl.bsky.social · 19/08/2026
Today I learned that it's easier to read a Nature paper on diffusion models, than the AI-Generated/Augmented article that's explaining the contents of the paper. Not faster but easier.
001
Andrea Grassi @andreasdl.bsky.social · 18/08/2026
At WCUS, WordPress.com just announced this beautiful initiative that I want to share, because I believe supporting education and students is the key for the future, and being able to experiment and create, but also to own their content. Here's why I think we need this
100
Andrea Grassi @andreasdl.bsky.social · 17/08/2026
"What are your hobbies?" Having this answer ready sometimes it's daunting and it carries a lot of judgment. But it shouldn't be. Hobbies evolve, they're not who we are, they're a part of who we'll be.
givemethechills.com
The Evolution of Hobbies
This beautiful video reminded me of how our hobbies change through life, how some hobbies are actually not hobbies, and how you evolve through the line of experiences that make you who you are today. I remember when I joined Automattic and I started just to talk to people, and they would ask me, “What’s your hobby?” And I didn’t have the answer.
000
Andrea Grassi @andreasdl.bsky.social · 17/08/2026
Every AI claim will sound confident. Before you'd only talk to a small number of highly confident people. Now everyone can be (and increasingly is). Yet confidence doesn't equate reality. Act accordingly.
000
Andrea Grassi @andreasdl.bsky.social · 17/08/2026
“Let me check that” Working with ChatGPT voice is fantastic, but you get some Memento Mori vibes whenever you ask something around what it just said and it replies with “let me check that” because it doesn’t have “memory”, but it needs to retrieve it.
010
Andrea Grassi @andreasdl.bsky.social · 15/08/2026
Profound and high quality thoughts and video. Absolutely worth it. www.youtube.com/watch?v=NFm2...
youtube.com
"Got any hobbies?"
YouTube video by Luda
000
Andrea Grassi @andreasdl.bsky.social · 13/08/2026
Here's what it means to work in the open nomad.blog/2026/08/04/u... Anne is leading the WordPress release and documenting the hard decisions and showing what it takes to lead such big project, all in the open. Lots to learn in each post.
nomad.blog
Using WordPress Studio to speed up testing 7.1
Before you read further, let this be an ask from yours truly to help test WordPress 7.1. Pick one thing! Get on a roll and maybe pick two! Any and everything helps. If you don’t have time to …
000
Andrea Grassi @andreasdl.bsky.social · 12/08/2026
Reminder that agents can read hidden text in images. Prompt injections like that might become more common. They are easily caught by any auto approval mode from frontier models, but local models might not be as good in detecting those attacks. Keep an eye on that.
000
Andrea Grassi @andreasdl.bsky.social · 11/08/2026
Didn't expect an error to lead to some learning, but here it is. I decided to investigate how come some code we wrote (for the same bug) was almost identical, and here's what I got out of it.
givemethechills.com
Getting Predictable AI Output
Recently me and a coworker had the unusual scenario where we were working on exactly the same bug. He was using Claude, and I was using Codex, but what was fascinating was the PR result: The code output was incredibly similar. It was a small change, not a big refactoring, but they were so close, almost identical. Different Models, right?
000
Andrea Grassi @andreasdl.bsky.social · 10/08/2026
Nvidia released a new Full-Duplex model called VoiceChat 11B. It takes 450ms to reply to you and processes the information while you speak. The cool part is that it has tool calling too. This is neat because this allows devs to build on this technology for different use cases.
100
Andrea Grassi @andreasdl.bsky.social · 06/08/2026
Reminder: The best code is the one you choose to not write. Here’s why
100
Andrea Grassi @andreasdl.bsky.social · 05/08/2026
Everyone is raving about text models, but I'm even more excited about vision, audio, and image models. They'll push creativity and accessibility forward, making AI truly open for everyone. GPT Voice can shift the way you work. Vision and image models might change the rest.
100
Andrea Grassi @andreasdl.bsky.social · 04/08/2026
I bought something unique and it took me 20 years to complete the checkout funnel. But it was worth it.
givemethechills.com
The Longest Checkout Flow: 20 Years
I recently bought a special kind of art that I discovered long time ago: a glass sphere with a unique depth effect when you look at it and when the light shines on it. I think it took me around 15 to 20 years to buy it. I first came across these spheres around 20 years ago, probably on Reddit. I do not remember exactly where I first encountered it.
000
Andrea Grassi @andreasdl.bsky.social · 04/08/2026
Working with agents using voice mode (specifically Codex voice) feels like a totally different way of interacting. It'll even feel more magical if we'll ever reach GPT 5.6-Luna levels of speed, with Sol quality
100
Andrea Grassi @andreasdl.bsky.social · 28/07/2026
Life for me has always been about scents, about things that linger and fade away, only to reappear later, evoked by the scents themselves.
givemethechills.com
Scents of Life
Life for me has always been made of scents. Of smells that emerge and unsettle you. Of thoughts that cling to a single fragrance molecule, to a particular accord, lingering not only on your clothes but in your memory. Like the scent of me embracing my mother on a cold winter night, when she came home from work and her coat carried the faint smell of smoke, along with the crisp note of the winter air woven into the worn leather of her jacket.
000
Andrea Grassi @andreasdl.bsky.social · 22/07/2026
Read this 3 times because the HuggingFace / OpenAI security incident is scary and fascinating at so many levels 👇
100
Andrea Grassi @andreasdl.bsky.social · 17/07/2026
Traits, Teaching, Time, Admiration, Credit, AI, Endings. Here's what it took me (and my recipe) to ask for guidance in today's world.
givemethechills.com
Asking for Guidance
How do you ask for guidance in today's world? Here's what comes to mind for me, based on a recent experience. Traits I wanted to ask for specific guidance and feedback from some people I admired, people I worked with. I wanted to understand if and how my work, and the way I am, impacted them. There are people who are able to spot their strong traits and weak traits easily.
000
Andrea Grassi @andreasdl.bsky.social · 16/07/2026
Here's an informative read if you're into prompt injections and want to understand the attack surface of them. OpenAI shared an article around GPT-Red, an automated way to test how their models perform against prompt injection.
100
Andrea Grassi @andreasdl.bsky.social · 15/07/2026
You should own your content, and you should be free to move it to whatever platform you want. Sadly, that is not always the case. This film from @codeforthepeople.com is a reminder of why people stand for open source and their values. Worth the 20' watch, trust me. www.youtube.com/watch?v=8lQi...
youtube.com
Code for the People | A Film by Bao Nguyen | Full Documentary
YouTube video by Code for the People
111
Andrea Grassi @andreasdl.bsky.social · 14/07/2026
3+1 books worth reading to think better about numbers and biases: 1️⃣ Factfulness, by Hans Rosling, Ola Rosling, and Anna Rosling Rönnlund: Instincts and biases tricks us, this will help you reading the world (with some optimism).
100
Andrea Grassi @andreasdl.bsky.social · 13/07/2026
Your agent can use MCPs to control apps and websites, yet sometimes you don't even know what they're doing. It shouldn't be like this. Security should always come first. @wpvip.com just released the Secure MCP, and I hope others will follow this lead, honestly, because it's just that useful.
110
Andrea Grassi @andreasdl.bsky.social · 08/07/2026
I didn't expect to write this, but here we are: Slop is actually good for you.
givemethechills.com
Slop is Inevitable, Slop is Good
More than 15 years ago, I decided I wanted to learn singing. Up until that point, I was the guy who didn't sing. I didn't sing in public, nor privately.We have this image of people singing in the shower. I didn't sing in the shower. Singing was not part of my radar, it was invisible to me, it was a skill I never thought existed, although I could clearly see (and hear) singers.
000
Andrea Grassi @andreasdl.bsky.social · 07/07/2026
(effective) Agentic Loops are the Deep Work state for LLM. What is your human loop, then?
000
Andrea Grassi @andreasdl.bsky.social · 07/07/2026
I always trusted numbers, and while I was often skeptical of stats, only recently I realized how easy it is to game them. The Will Rogers phenomenon changed the way I see the world.
givemethechills.com
The Problem with Statistics and Averages: The Will Rogers Phenomenon
One of the problems with statistics and numbers is that we trust them as if they're always trustworthy. But the problem with those numbers is that they depend on which domains and which sources they're using. Averages are even trickier than that. One concept that blew my mind recently is the Will Rogers phenomenon. A Simple Example Let's say you're a sales manager.
000
Andrea Grassi @andreasdl.bsky.social · 06/07/2026
There's a new discussion paper based on 26.000 Chinese students comparing those using AI vs those not using it. The result: AI kids were faster in the short term, got worse after 2 years. No surprise. We're still adapting. We can't just replace our skills with AI and be done with it.
100
Andrea Grassi @andreasdl.bsky.social · 02/07/2026
Either you're locked into a system, or you're free to change your mind, with no chains. Having something that stands for that freedom is truly priceless in today's world, where most of the products try to lock you in. Be brave (and read this beautiful post from @zeldman.bsky.social ).
132
Andrea Grassi @andreasdl.bsky.social · 01/07/2026
Imagine a self evolving SAAS software where you can pay the tokens to build a feature you needed, in addition to a monthly fee. If others will use it then I the fee you pay will be detracted by your monthly costs. If you’ll be the only one using it, that’s it. Would you use it?
010
Andrea Grassi @andreasdl.bsky.social · 30/06/2026
Multiple Dimensions of AI Cerebras just announced Gemma 4 31B running at over 1,800 tokens per second.
100
Andrea Grassi @andreasdl.bsky.social · 25/06/2026
This resonates hard with me. It's not happening every day, but it is happening. It's hard to say if it's caused by willpower depletion or just exhaustion. Not so unexpectedly, @elijahpotter.dev added a very smart idea at the end of the blog post. Worth a read. elijahpotter.dev/articles/slo...
elijahpotter.dev
Slop Paralysis
A complete or partial loss of function while reviewing the output of a coding agent.
010
Andrea Grassi @andreasdl.bsky.social · 24/06/2026
OpenAI just announced its first chip, and it’s no surprise. AI is not cheap. It takes quite a lot of money and electricity to run, and there’s a reason companies are starting to build chips.
100
Andrea Grassi @andreasdl.bsky.social · 23/06/2026
To understand we need to let go of our ego and stop trying to ask smart questions. Simple, basic questions are incredibly effective to get deeper into any problem and get a superior understanding. Letting the ego go is the hard part.
givemethechills.com
Understanding by Asking Questions
One of the things that makes understanding easier is asking questions. We are often tricked by our own intelligence because we think we understand, whereas in fact we don’t: We’re not truly understanding something. What’s stopping us from understanding the solution is getting deeper into the problem.But how do you get much deeper into the problem? It’s something that I also touched…
000
Andrea Grassi @andreasdl.bsky.social · 22/06/2026
ChatGPT is offsetting Pulse in favor of Scheduled actions and this is a strong reminder of what AI truly unlocked for product development: The acceptance of experimental features.
100
Andrea Grassi @andreasdl.bsky.social · 19/06/2026
AI will happily spend tokens all day long. It won’t question assumptions. It won’t ask whether the idea is worth pursuing. Which brings us back to the same product question we’ve always had: What’s the value of what you want to build?
givemethechills.com
Cutting Costs / Adding Value
What is price? It can be a measure to "weight value".For example, there was this person on Twitter that stated stated how $30 per PR shipped is a good price, and I tend to agree. The thing we have with AI now is that you can spin up a model ask it to build something. Guess what? The model will agree and will start using tokens (spending tokens).
110
Andrea Grassi @andreasdl.bsky.social · 18/06/2026
The Midjourney pivot to medical stuff is mindblowing. I'm really curious to see how that will turn out.
000
Andrea Grassi @andreasdl.bsky.social · 17/06/2026
“60% say AI in a brand’s messaging is a turnoff, not a feature.” That’s one of the findings from the @wpvip.com article "Future of the Web 2026: AI Brand Visibility Research." And… I kind of agree. I mean, who really wants to talk to an AI unintentionally? (continues)
111