The Wiert Corner – irregular stream of stuff

Jeroen W. Pluimers on .NET, C#, Delphi, databases, and personal interests

  • My badges

  • Twitter Updates

  • My Flickr Stream

  • Pages

  • All categories

  • Enter your email address to subscribe to this blog and receive notifications of new posts by email.

    Join 1,860 other subscribers

Archive for the ‘GitHub Copilot’ Category

GitHub parody poster (but actually truth): “We trained copilot on your code. That’s why it sucks”

Posted by jpluimers on 2025/06/18

The image [Wayback/Archive] GjXf0EPWEAAaAoz.jpg:orig (1528×2048) posted by [WaybackSave/Archive] Deepak Kumar (दीपक) 😷💉 on X: “@Grady_Booch @jsngr” was a response to [Wayback/Archive] Grady Booch on X: “@jsngr …”

«

And what to we call a language with sufficient expressiveness and precision able to construct executable artifacts?

We call it a “programming language”.

At best, AI will revolutionize software development on the same manner as did the invention of compilers. The entire history

»

which also spared these interesting responses:

  1. [WaybackSave/Archive] Adiva 🏳️‍⚧️ on X: “@Grady_Booch @jsngr Also: libraries/packages. It’s been ages since anybody has needed to write every single thing by themselves. Lots of software writing is already creating logic that connects more specialized code written by other people without having understand its internals fully.”
  2. [WaybackSave/Archive] Fast Magpie on X: “@Grady_Booch @jsngr This is the correct take. The number of software engineers required to produce software will decline. The breadth and depth of skill required of those engineers will likely increase. The overall demand for software will increase.”

The library problem is even bigger than the XKCD single dependency one, as there are so many of those fragile dependencies which I discussed in Every conversation about dependencies since 2020 uses the same XKCD 2347 based image, which is a problem on multiple levels. AI will introduce even more dependencies and likely many of them from unknown background causing huge copyright issues.

Image (full size below) from [Wayback/Archive] Tweet JSON

--jeroen


Posted in AI and ML; Artificial Intelligence & Machine Learning, Conference Topics, Conferences, Development, Event, GitHub Copilot, LLM, Software Development | Leave a Comment »

According to Imgur: The new abbreviations for Generative AI is PISS

Posted by jpluimers on 2024/10/29

Plagiarized Information Synthesis System as per [Wayback/Archive] Spot on! AI can piss off. – electronics post – Imgur

Read the rest of this entry »

Posted in AI and ML; Artificial Intelligence & Machine Learning, ChatGPT, Development, Fun, GitHub Copilot, GPT-3, GPT-4, LLM, Meme, Software Development | Leave a Comment »

Kevlin Henney on generative AI creating job security for programmers:

Posted by jpluimers on 2024/08/13

Kevlin Henney being interviewd by Richard Seidl

Kevlin Henney being interviewd by Richard Seidl [Wayback/Archive] MDVxFQrqZnh1OxlP.jpg (1200×675)

The quote from this abstract of the January 2024 interview with Kevlin Henney by Richard Seidl  is important:

You really need to understand history. First of all, you need to understand history. Then, you need to understand language. And you need to go and talk to some customers. And then, you will realize how safe your job is. Because programming is not merely the assembly of syntax. It is the application of precision. It is the seeking of precision.And what is the answer? What is it that I’m trying to do?And it turns out that if you specify something badly in natural language, it works out even worse than if you did it in code.And we already know, for example– we can actually take inspiration from the most widely used programming paradigm on the planet, the spreadsheet. What we know from the spreadsheet is that most people who use a spreadsheet do not have a software development background.

Yes.

We also know that most spreadsheets are unmaintainable, incomprehensible, and buggy. If we are saying that the future of software development is people who are not software experts doing this stuff, your job is safe.

It is a fragment of the vodcast episode [Wayback/Archive] Software Engineering im Jahr 2034 – Richard Seidl which limits the quote to this:

Read the rest of this entry »

Posted in AI and ML; Artificial Intelligence & Machine Learning, Conference Topics, Conferences, Development, EKON, Event, GitHub Copilot, LLM, Software Development | Leave a Comment »

“Oh shit git” seems to have been succeeded by “Oh shit GitHub Copilot”: ‘Downward Pressure on Code Quality’

Posted by jpluimers on 2024/01/29

Not sure about you, but when I write code I want it to be better – way beter even – than average code.

The problem with any LLM based Generative AI is that it generates text based on the average of the past corpus they were trained with at the time they were trained.

It is exactly why I have been advocating for a while: be careful when using Generative AI, as you get generated text based on the combination of averaging over the LLM corpus with the relatively small prompt you phased trying to reflect a tiny bit of the model of the reality you are trying to write software for.

So I was not at all surprised by this article: [Wayback/Archive] New GitHub Copilot Research Finds ‘Downward Pressure on Code Quality’ — Visual Studio Magazine.

Read the rest of this entry »

Posted in AI and ML; Artificial Intelligence & Machine Learning, Development, GitHub Copilot, LLM, Software Development | Leave a Comment »