3 ms·

GTP is interesting enough without people pretending that it is more interesting than it is. This "new programming language" doesn't look new at all. It looks li

GTP is interesting enough without people pretending that it is more interesting than it is. This "new programming language" doesn't look new at all. It looks like the same old thing.

marviel 4y ago
You should read further down in the article where it expands on some more advanced concepts -- but in general you're right -- it is really similar to existing languages. I wonder as its power changes, what paradigms it might suggest we break out of. I'd also recommend reading its responses that are non-code in detail -- for instance, in the comment where it introduces TenetLang, GPT-4 actually calls out that it's re-using ideas. I would assume intentionally, because it knows what people are most familiar with. > GPT-4 > In this new language, called "TenetLang," we'll combine a simple syntax inspired by Python with some features from functional and object-oriented languages. Here is an overview of some design choices: - Indentation-based scoping, similar to Python. - Strong typing with type inference, inspired by TypeScript and Kotlin. - First-class functions and closures, similar to JavaScript. - Immutable data structures by default, with optional mutable counterparts. - A concise lambda syntax for anonymous functions. - Pattern matching and destructuring, inspired by Haskell and Rust. - Built-in support for concurrency using async/await and lightweight threads. - Interoperability with other languages using a Foreign Function Interface (FFI).
- karmakaze 4y ago
  GPT-4 didn't learn the distinction between strong types and static typing. That makes sense as much writing also doesn't make the distinction. e.g. Python has strong types, can't add 1 + "2".
vidarh 4y ago
That is true for most "new" human produced.programming languages too. The pool of truly novel programming languages is extremely shallow. Even the number of languages bringing anything new to the table is very limited. So, it doesn't really tell us much (either direction)
- midoridensha 4y ago
  Unlike languages invented in the 1960s, today's new languages seem to basically recycle ideas from existing languages while attempting to fix the perceived deficiencies in those languages. This shouldn't be a surprise; any time a new industry arises, most of the truly groundbreaking and novel inventions will be in the beginning, when there's no experience to draw on. Now that programming has been around for the better part of a century, we know a lot about what works and what doesn't.

GTP is interesting enough without people pretending that it is more interesting than it is. This "new programming language" doesn't look new at all. It looks like the same old thing.

marviel 4y ago

You should read further down in the article where it expands on some more advanced concepts -- but in general you're right -- it is really similar to existing languages. I wonder as its power changes, what paradigms it might suggest we break out of. I'd also recommend reading its responses that are non-code in detail -- for instance, in the comment where it introduces TenetLang, GPT-4 actually calls out that it's re-using ideas. I would assume intentionally, because it knows what people are most familiar with. > GPT-4 > In this new language, called "TenetLang," we'll combine a simple syntax inspired by Python with some features from functional and object-oriented languages. Here is an overview of some design choices: - Indentation-based scoping, similar to Python. - Strong typing with type inference, inspired by TypeScript and Kotlin. - First-class functions and closures, similar to JavaScript. - Immutable data structures by default, with optional mutable counterparts. - A concise lambda syntax for anonymous functions. - Pattern matching and destructuring, inspired by Haskell and Rust. - Built-in support for concurrency using async/await and lightweight threads. - Interoperability with other languages using a Foreign Function Interface (FFI).

karmakaze 4y ago
GPT-4 didn't learn the distinction between strong types and static typing. That makes sense as much writing also doesn't make the distinction. e.g. Python has strong types, can't add 1 + "2".

vidarh 4y ago

That is true for most "new" human produced.programming languages too. The pool of truly novel programming languages is extremely shallow. Even the number of languages bringing anything new to the table is very limited. So, it doesn't really tell us much (either direction)

midoridensha 4y ago
Unlike languages invented in the 1960s, today's new languages seem to basically recycle ideas from existing languages while attempting to fix the perceived deficiencies in those languages. This shouldn't be a surprise; any time a new industry arises, most of the truly groundbreaking and novel inventions will be in the beginning, when there's no experience to draw on. Now that programming has been around for the better part of a century, we know a lot about what works and what doesn't.