Re: Discussion: Using LLMs the Right Way: 10/1/2025 7pm Eastern Daylight time
Deirdre Saoirse Moen <[email protected]> Wed, 1 Oct 2025 16:06:08 -0700
| Newsgroups | gmane.org.user-groups.linux.cabal |
|---|---|
| Message-ID | <[email protected]> |
--===============1990168223== Content-Type: multipart/alternative; boundary="Apple-Mail=_63971F33-4CB2-481E-B3C8-F0A5558067F0" --Apple-Mail=_63971F33-4CB2-481E-B3C8-F0A5558067F0 Content-Transfer-Encoding: quoted-printable Content-Type: text/plain; charset=utf-8 On Oct 1, 2025, at 15:59, Don Marti <[email protected]> wrote: >=20 > I have a feeling that we're going to get to something similar with = LLMs. The tricks that people are coming up with in order to avoid = introducing severe breakage because of LLMs are also going to help = improve software quality in general, and the actual LLM will turn out to = be like the pumpkin in the pumpkin spice. Yes, in fact the single greatest benefit of the LLM boom may be that = code bases are being refactored to be simple enough so LLMs can = understand them, which also has the side benefit of making them easier = for human devs to understand them. While I criticized this one LLM example within $PRIOR_FIRM, some of the = suggestions the dev LLM made were helpful, and it reminded me of things = that I=E2=80=99d have otherwise discovered only after going through a = compile cycle. It was also very good generally at pointing me to where = to look, which saved time. Except this once when it happened to cost me = about two hours of work. Deirdre= --Apple-Mail=_63971F33-4CB2-481E-B3C8-F0A5558067F0 Content-Transfer-Encoding: quoted-printable Content-Type: text/html; charset=utf-8 <html><head><meta http-equiv=3D"content-type" content=3D"text/html; = charset=3Dutf-8"></head><body style=3D"overflow-wrap: break-word; = -webkit-nbsp-mode: space; line-break: after-white-space;">On Oct 1, = 2025, at 15:59, Don Marti <[email protected]> = wrote:<br><div><blockquote type=3D"cite"><br = class=3D"Apple-interchange-newline"><div><span style=3D"caret-color: = rgb(0, 0, 0); font-family: LucidaGrande; font-size: 14px; font-style: = normal; font-variant-caps: normal; font-weight: 400; letter-spacing: = normal; text-align: start; text-indent: 0px; text-transform: none; = white-space: normal; word-spacing: 0px; -webkit-text-stroke-width: 0px; = text-decoration: none; float: none; display: inline !important;">I have = a feeling that we're going to get to something similar with LLMs. The = tricks that people are coming up with in order to avoid introducing = severe breakage because of LLMs are also going to help improve software = quality in general, and the actual LLM will turn out to be like the = pumpkin in the pumpkin spice.</span><br style=3D"caret-color: rgb(0, 0, = 0); font-family: LucidaGrande; font-size: 14px; font-style: normal; = font-variant-caps: normal; font-weight: 400; letter-spacing: normal; = text-align: start; text-indent: 0px; text-transform: none; white-space: = normal; word-spacing: 0px; -webkit-text-stroke-width: 0px; = text-decoration: none;"></div></blockquote></div><br><div>Yes, in fact = the single greatest benefit of the LLM boom may be that code bases are = being refactored to be simple enough so LLMs can understand them, which = also has the side benefit of making them easier for human devs to = understand them.</div><div><br></div><div>While I criticized this one = LLM example within $PRIOR_FIRM, some of the suggestions the dev LLM made = were helpful, and it reminded me of things that I=E2=80=99d have = otherwise discovered only after going through a compile cycle. It was = also very good generally at pointing me to where to look, which saved = time. Except this once when it happened to cost me about two hours of = work.</div><div><br></div><div>Deirdre</div></body></html>= --Apple-Mail=_63971F33-4CB2-481E-B3C8-F0A5558067F0-- --===============1990168223== Content-Type: text/plain; charset="us-ascii" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit Content-Disposition: inline _______________________________________________ conspire mailing list [email protected] http://linuxmafia.com/mailman/listinfo/conspire --===============1990168223==--