ghuntley

15 ms·

241.
Coding Is Over. Now What?(svpino.com)
6 points by ghuntley 2y ago|2 comments
242.
A Model Context Protocol Server (MCP) for Microsoft Paint(ghuntley.com)
1 points by ghuntley 2y ago|0 comments
243.
Nix that looks like Bazel(fzakaria.com)
4 points by ghuntley 2y ago|0 comments
244.
MCP Security Notification: Tool Poisoning Attacks(invariantlabs.ai)
1 points by ghuntley 2y ago|0 comments
245.
Everything is Ghibli(carly.substack.com)
213 points by ghuntley 2y ago|340 comments
246.
Great AI Steals(world.hey.com)
2 points by ghuntley 2y ago|0 comments
247.
AbletonMCP – Ableton Live Model Context Protocol Integration(github.com)
57 points by ghuntley 2y ago|11 comments
248.
AI Model Context Protocol (MCP) and Security(community.cisco.com)
6 points by ghuntley 2y ago|0 comments
249.
Time between disengagements: the rise of the software conductor(gitpod.io)
2 points by ghuntley 2y ago|0 comments
250.
by ghuntley 2y ago
I have been able to create a brand new programming language purely based on specs and the eval loop technique. If I can build an ESO language using a foundational model (ie. no pre-existing training data) and program with it, then Clojure i
251.
by ghuntley 2y ago
I actually agree with this. This will be true, as long as Rust doesn’t break their language semantics. Zig keeps making breaking changes and when they do it essentially means all the pre-trained data is incorrect. Going forward it is incred
252.
by ghuntley 2y ago
Nailed it. These LLMs do not understand code. They are silly string lookup services. Compilers with soundness tend to be INCREDIBLY verbose with their error messages which these LLMs use to ingest to get better outcomes on the re-attempt…
253.
by ghuntley 2y ago
I’ve stumbled across folks using TLA+ / LEAN / COQ. Instead of ramming TypeScript/Whatever into the context window, they just stuff the formal proofs into the context window. Then code generate off em… That’s most Chad patter
254.
by ghuntley 2y ago
Hard disagree. The best language for AI is one that has strong compiler soundness (ie type system) that can reject incorrectness and provide feedback as to correctness to refactors. One can substitute with property based tests but it won’t
255.
AI for my 10-year-old son(ghuntley.com)
3 points by ghuntley 2y ago|1 comments
256.
by ghuntley 2y ago
It's not the LLM but the current generation of coding tools that need to be adjusted/tweaked to deal with em, yes.
257.
by ghuntley 2y ago
<3 you get it. It's a new skillset to be learned and when it clicks, you get incredible outcomes!
258.
by ghuntley 2y ago
> Lots of people don't like to hear this, but if you are not using AI today in some capacity, or your company is lagging in its adoption your career is at risk, especially if you are still relatively young. This said as somebody who
259.
by ghuntley 2y ago
> I'm wondering if the LLM considers "code" and "spec" as two separate programming languages, or one as a programming, and the other as a human language? Not sure if it makes a difference or not for its internal
260.
by ghuntley 2y ago
It is sound! The real context window is indeed smaller than the advertised marketing number. This gets interesting in verbose languages such as Java, where I've seen +16k tokens wasted on the classpath/import stanzas within a clas
261.
by ghuntley 2y ago
Here https://github.com/DanielJoyce/ls-rs is some-one who applied both techniques to convert `ls` to `rust` via an `ls.objdump`
262.
by ghuntley 2y ago
> manages things like naming conventions, APIs from other apps, etc. These particular concerns can be handled via https://ghuntley.com/stdlib and it works very very well.
263.
by ghuntley 2y ago
That's an interesting question, and I spend plenty of time exploring this at my current client. The best tip I can give you right now is really - it depends. For example .NET/Java have this horrendous convention of splitting files
264.
by ghuntley 2y ago
oh interesting. One thing about the specy, is it's incredibly hard to hook in the eval technique (see middle/bottom of https://ghuntley.com/specs/ ) as deployment is manual/human. So I had to drive it al
265.
by ghuntley 2y ago
covered here https://news.ycombinator.com/item?id=43387590 but yes, indeed you still need to watch it like a hawk when using it outside of a tight eval loop (ie. make all | make property-tests) and/or implement some ot
266.
by ghuntley 2y ago
Yeah, this was pointed out to me after publishing. Specifically, when the ASM was converted over /specs, it incorrectly wrote the amount of sales tax to the specs. Thus, the z80 implementation took that incorrect /spec and made it
267.
by ghuntley 2y ago
It was more about applying /specs, to generate specs. Then applying /stdlib + /specs to generate an application for another platform without using the original reference application/disassembler. Proving a point that the
268.
by ghuntley 2y ago
Give it a try. You'll be surprised at what can be achieved when the LLM is driven via /specs (business requirements) + /stdlib (steering LLM technical outcomes). The end result, when driven by a good eval loop (property-based
269.
Converting C to ASM to specs and then to a working Z/80 Speccy tape(ghuntley.com)
80 points by ghuntley 2y ago|73 comments
270.
by ghuntley 2y ago
It wouldn't have. Without him, there would be no WikiLeaks et al.

More ›