Sal. 🍷

Agile. Technology. Management.

[Sal. 🍷 @saieva]

Agile is not fast, it's frequent. https://t.co/Xjbi9HcSRj

[Sal. 🍷 @saieva]

@stephen_wolfram God bless Elise. 🙏🙏🙏

[Sal. 🍷 @saieva]

@KentBeck The LLMs need the practices of the craft now, so the lessons are not lost nor are they obsolete. Humans still need knowledge of the craft to assess the results.

Work iteratively. Deliver incrementally. - Those lessons of how to work remain priceless.

[Sal. 🍷 @saieva]

@unclebobmartin @BucKuc30673 I am a subscriber of the full Clean Code and many other @thecleancoders series. Agentic Discipline has been good so far but there is something missing. As enjoyable as the glitz was, it's not important. We need something like the Java Case Study series for agentic development.

[Sal. 🍷 @saieva]

New interface ideas for desktops, such as touch-screen, haven't been great. This seems like the way.

https://t.co/RUU6kr4bae

[Sal. 🍷 @saieva]

@JoannaStern Not an email client, but for a completely new take on email checkout @heyhey

[Node.js @nodejs]

Are you currently hiring for a role that includes using Node.js? Reply with a link to the opening and any relevant context.

If you're not, we'd appreciate a repost for visibility 💚

[Sal. 🍷 @saieva]

@unclebobmartin Amazing!

[Sal. 🍷 @saieva]

@unclebobmartin I hope the AI Coding Product Managers (or their equivalents) at Anthropic, OpenAI, and Google are following your experiences and feedback here.

[Sal. 🍷 @saieva]

@unclebobmartin Karpathy recommended keeping context windows as small as possible by frequently starting new chats. Could that help with the coding CLIs?

How I use LLMs by Andrej Karpathy
https://t.co/ALfUwokx7s

[Sal. 🍷 @saieva]

@unclebobmartin Maybe, if humans never need to read the code. We don't read the assembly output of compilers.

But if we do need to read the code, and if the AI is truly better than humans, then the AI should be able to generate and maintain highly readable code.

[Sal. 🍷 @saieva]

@markgurman Apple invested in Maps at a time when there were other choices (ie, Google and TomTom) because, sensibly, they needed to own that IP. It's confusing they wouldn't similarly want to own their Foundation Model, LMM, and ChatBot.

[Sal. 🍷 @saieva]

"Putting the personal connection first doesn’t just make collaboration possible. It makes it enduring." Don’t Underestimate the Value of Professional Friendships

https://t.co/0MJ086YWJv

[Sal. 🍷 @saieva]

@MKBHD The Gemini deal with Google may demonstrate how late, or disoriented, Apple is with AI. Long term they will need their own IP. That's always been their strategy.

[Sal. 🍷 @saieva]

@MKBHD When compared to Maps, it's confusing that Apple is not building their own LLM/AI. They made a massive investment in Maps when there were other drop-in app choices at the time (eg, Google Maps and TomTom). How is having a proprietary Foundation Model less important than Maps?

[Sal. 🍷 @saieva]

@lexfridman When an LLM responds to a question, how does it know when its response is complete? How does it know when to stop?

[Sal. 🍷 @saieva]

@unclebobmartin I haven't used an LLM like this yet. Does Grok have access to the IDE to make changes and run tests?

[Sal. 🍷 @saieva]

"Your job as a leader isn’t to be the smartest person in the room; it’s to make the room smarter than any one person." Great Leaders Empower Strategic Decision-Making Across the Organization.

https://t.co/ZGUjTst5XX

[Sal. 🍷 @saieva]

Amazon CTO Werner Vogels publishes the seven laws of Frugal Architecture.

https://t.co/2XtLsoCdDD

[Sal. 🍷 @saieva]

"The result is what we call a black box: a system whose outputs we can observe, but whose inner workings remain opaque." Using AI to reverse engineer legacy enterprise systems.

https://t.co/7I2QXSzhhQ

[Lex Fridman @lexfridman]

Programming with AI is insanely fun. Process is:
1. generate code
2. read & understand code that was generated
3. make small changes "manually" (still with great autocomplete)
4. test & debug
5. make big changes with new prompt
6. go back to step 1

Pure vibe coding skips step 2 & 3. And I think we'll need human expertise & experience for steps 2, 3 (and 4) for quite a while.

But holy shit, I'm learning much faster, being way more productive, and having more fun.

Not sure we're close to "AGI/ASI", but the software engineering world is definitely getting transformed very rapidly. It feels surreal to be experiencing it directly on a daily basis. Of course, there are both scary (jobs, security) & exciting (productivity, access) consequences to this transformation, as with all powerful technology.

This post was fully written by human in one-shot without spellcheck, it's 100% organic human writing 🤣