Sunena Mirza

It is a question of choosing between two evils for us now. Neither option is completely free of flaws.

Human: Recruiters with "gut feelings" who harbor unconscious bias. they reject excellent candidates who just didn't go to the "right" school or didn't just "click." Inconsistent, unfair, and un-auditable.

AI: Algorithms whose training datasets are themselves replete with historical biases. They increase the scale of discrimination at light speed, becoming so-called black boxes that end up rejecting qualified candidates for reasons that humans cannot even fathom.

We are truly deciding to exchange messy, subjective human prejudice for cold, ruthlessly efficient algorithmic prejudice. Is that really an upgrade?

I have been vibe-coding exclusively for about a month now. I am a senior software engineer and I have always been interested in the "no coding" approach to software development. I want software development to be easier, more accessible to everyone. So I switched to vibe-coding daily and writing less code directly or with AI assistance. When I read about other's experience, I feel there is this notion that vibe coding means the computer will think through all the details of the software and deliver an error-free product. This is not true and sets us up for frustration.
Vibe coding can be immensely powerful if we are willing to do some ground research. Start with Claude Code best practices. You do not have to understand all the technical concepts right away; just an overview may help a lot. And many of the points can apply if you are using tools other than Claude Code.

Majority of software for our general use has a UI layer and a data + controller layer. The UI layer is what we typically call "frontend" and the data + controller layer "backend". The UI layer loads in our browsers, desktop or mobile. We can even have native mobile apps for it. The backend layer stays on a computer connected to the Internet that we access from our frontend. We call the Internet connected computer (that serves the data for frontend) a "server". The backend and frontend communicate using an API (application programming interface). When we combine a software with all three parts and the tools needed to put everything in their place, we generally call it "full-stack software".
With these basic terms (frontend, backend, API) we can build simple, usable software much more easily than we may think. We can use OpenAI, Claude, Gemini or any of the top open source LLMs (large language model) to help us break down our software into parts that then can be coded by Claude Code or a similar code generation tool.
There will be obstacles, there will be confusing steps and frustration. That happens when we send our software ideas to developers too. What I want to encourage is to try. Learn about the topics a little, some basics of a software development workflow, how GitHub manages software development, what does DigitalOcean or Amazon Web Services provide, etc. It is OK if we do not get a fully working, bug-free software launched in one go. If we can learn how to use these new tools, so many more people can build the software they need.

A few weeks ago, @Veo emerged.

A few days ago, I tried @Higgsfield. The output is here.

Yesterday, Chris hunted @Midjourney Midjourney V1 Video Model.

And today I read that @Perplexity has a video generation available on Ask Perplexity!