I finally updated my portfolio! 🥳 I'd been using the same version since 2023 🤣 Some of the text still needs work, but yeah, it was definitely time for an update! Also, I finally bought a domain 😅 Here's a little before and after 👇 cmglezp.dev

Very good UX. Looks like they spent all their credits on Twilio 😅

The amount of hate I've seen here in the last two days, for free, is unreal.

gpt-6.1 (and I don't know about other versions) love to write code without line breaks. the code looks very ugly and cluttered.

All my time can be summed up as: - thinking about how to solve problems and iterating on solutions with AI. - reading AI-generated code. - reading my coworkers' code (also AI-generated code).

AWS launched Lambda Web Functions on October 12, bringing a native HTTP server model to Lambda. Traditional Lambdas are now called Event Functions. Web Functions let you run a Node.js backend using Express, Hono, Fastify or NestJS without event-to-HTTP adapters. AWS provides HTTPS, manages scaling and scales the application to zero when idle. Multiple concurrent requests can share one environment, reusing the initialized application and connection pools. While one request waits for a database, the server can handle others. This makes Web Functions interesting for: - REST and GraphQL APIs with database or external-service calls. - Server-rendered applications, such as Next.js. - AI backends that stream responses to clients. - HTTP microservices with variable traffic. They also work with CloudFront for applications that combine static and dynamic content. CloudFront can cache static assets while Web Functions handles server rendering and API requests. Billing separates active CPU from memory consumption. Reference rates for US East, N. Virginia, in USD are: - $0.30 per million requests. - Approximately $0.0997 per active vCPU-hour. - Approximately $0.0132 per GB-hour of consumed memory. Waiting on I/O consumes little CPU, but memory remains billable while the environment exists, including between requests. Data transfer and other services add costs. AWS pricing. Cold starts still exist: your application must initialize whenever a new environment starts. The benefit is sharing that initialized server across concurrent requests. Announcement: aws.amazon.com/blogs/aws/ann… Pricing: aws.amazon.com/lambda/pricin…

People are really quick with these things haha.

why is Cursor doing this? are you trying to force me to use Grok 4.7?

after a month at a new job, I'm realizing how much damage the last seven months of heavy AI-assisted development did to my engineering skills. I used Claude Code almost every day (20x plan), and while it made me incredibly productive, I slowly lost a lot of low-level technical judgment along the way. I'm recovering those skills now, but it's honestly concerning to see how much AI can affect the way you think when you overuse it. I still believe AI is one of the most valuable tools we have. Without it, onboarding into large projects, understanding legacy code, tracing dependencies, and exploring unfamiliar codebases would take much longer. It helps me investigate problems faster and understand systems that would otherwise take days or weeks to learn. the problem isn't AI itself. The problem starts when AI is combined with a culture that only cares about shipping faster and faster. In that environment, there's no time to stop, think, analyze, or understand what you're building. Everything gets delegated to the AI because deadlines matter more than understanding. in the short term, the company benefits. Features get shipped quickly and productivity metrics look great. But in the long term, you may be the one paying the price. technical interviews, onboarding into new codebases, debugging difficult issues, system design discussions, and many day-to-day engineering tasks still require strong fundamentals and good technical judgment. If you've spent months outsourcing most of your thinking, you'll eventually notice the gap. the solution isn't "don't use AI." Use AI as much as possible. Let it accelerate your workflow. just make sure you're still exercising your own judgment and understanding the decisions being made. Otherwise, you might wake up one day and realize that while the company became more productive, you became less capable. these are just my thoughts after the last seven months. I'm curious to hear what others think.

Two major issues are missing here: Input validation: The user can literally pass anything. Also, this will probably throw a type error because amount is not being read from anywhere. Transaction & rollback: What happens if the debit succeeds but the credit fails? You would end up with an inconsistent state. The application should roll back the transaction if anything fails between these two operations. Both operations should be executed as a single unit of work and either complete successfully together or be rolled back together if either one fails.

I've used Github Copilot some days because I have it with my job, but wow, it's garbage. I don't know why exactly, but feels terrible. It consumes a lot of its "credits", it doesn't have the frontier models and the its router doesn't work well.

I've been using GLM 5.3 Flash for my main workflows, and wow, I really don't need more. Sometimes I use GPT 5.6 Sol for complex tasks or deep thinking but it's really impressive the performance of GLM. It's really cheap and very fast.