1. Forum
  2. Usenet
  3. COMP.MISC

  • People Are Googling Fake Sayings To See AI Overviews Explain Them

    From Lawrence D'Oliveiro@21:1/5 to All on Thu Apr 24 21:48:59 2025
    Here’s yet another way to have fun with AI, AI, everywhere. If Google
    wants to do AI analyses in response to your search queries, why not
    try giving it whimsical, not to say fictional, search queries, to do
    the analysis on?

    And true to form, the AI supercomplex does not disappoint <https://www.zdnet.com/article/people-are-googling-fake-sayings-to-see-ai-overviews-explain-them-and-its-hilarious/>.

    Try it yourself. I put in “you can't make avocado without breaking
    legs”, and it came back with some paragraphs that began

    The statement "you can't make avocado without breaking legs" is a
    humorous exaggeration, implying that preparing or consuming
    avocados is a challenging and dangerous task. In reality, avocados
    are not difficult to prepare, and their consumption does not pose
    any risk of breaking legs.

    followed by some genuine tips on how to prepare avocados, and ending with

    While there's a popular saying about avocado toast being a "foodie
    trend," the reality is that preparing and enjoying avocados is a
    simple and safe process. The statement about breaking legs is just
    a funny way to highlight the popularity and sometimes the
    perceived "effort" associated with enjoying this fruit.

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From Richmond@21:1/5 to Lawrence D'Oliveiro on Fri Apr 25 12:20:00 2025
    Lawrence D'Oliveiro <ldo@nz.invalid> writes:

    Try it yourself. I put in “you can't make avocado without breaking
    legs”, and it came back with some paragraphs that began

    Is this a case of rubbish in, rubbish out? or ask a silly question, get
    a silly answer.

    ChatGPT, by its own admission, holds up a mirror.

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From Richard Kettlewell@21:1/5 to Richmond on Fri Apr 25 15:52:58 2025
    Richmond <dnomhcir@gmx.com> writes:
    Lawrence D'Oliveiro <ldo@nz.invalid> writes:
    Try it yourself. I put in “you can't make avocado without breaking
    legs”, and it came back with some paragraphs that began

    Is this a case of rubbish in, rubbish out? or ask a silly question, get
    a silly answer.

    ChatGPT, by its own admission, holds up a mirror.

    I think a persistent bluffer is a good model for it. Sometimes right by
    luck, but very much not reliable for anything that matters.

    Most recently AI slop has created a new kind of supply chain attack:

    https://www.bleepingcomputer.com/news/security/ai-hallucinated-code-dependencies-become-new-supply-chain-risk/

    --
    https://www.greenend.org.uk/rjk/

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From Richmond@21:1/5 to Richard Kettlewell on Fri Apr 25 18:01:52 2025
    Richard Kettlewell <invalid@invalid.invalid> writes:

    Richmond <dnomhcir@gmx.com> writes: > Lawrence D'Oliveiro
    <ldo@nz.invalid> writes: >> Try it yourself. I put in “you can't make >avocado without breaking >> legs”, and it came back with some
    paragraphs that began

    Is this a case of rubbish in, rubbish out? or ask a silly question,
    get a silly answer.

    ChatGPT, by its own admission, holds up a mirror.

    I think a persistent bluffer is a good model for it. Sometimes right
    by luck, but very much not reliable for anything that matters.

    Most recently AI slop has created a new kind of supply chain attack:

    https://www.bleepingcomputer.com/news/security/ai-hallucinated-code-dependencies-become-new-supply-chain-risk/

    Well I guess they didn't have a code review. Even human beings get it
    wrong. I see GPT is well down the list.

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From Lawrence D'Oliveiro@21:1/5 to Richard Kettlewell on Fri Apr 25 23:40:08 2025
    On Fri, 25 Apr 2025 15:52:58 +0100, Richard Kettlewell wrote:

    Most recently AI slop has created a new kind of supply chain attack:

    https://www.bleepingcomputer.com/news/security/ai-hallucinated-code-dependencies-become-new-supply-chain-risk/

    ... aaand I’m sure somebody will come up with an AI-based solution to
    that.

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From Jim Jackson@21:1/5 to Lawrence D'Oliveiro on Mon Apr 28 12:42:26 2025
    On 2025-04-25, Lawrence D'Oliveiro <ldo@nz.invalid> wrote:
    On Fri, 25 Apr 2025 15:52:58 +0100, Richard Kettlewell wrote:

    Most recently AI slop has created a new kind of supply chain attack:

    https://www.bleepingcomputer.com/news/security/ai-hallucinated-code-dependencies-become-new-supply-chain-risk/

    ... aaand I???m sure somebody will come up with an AI-based solution to
    that.

    and 5-20% of such solutions will be be hallucinated solutions!

    Mind you, way back in the day I saw degree student's software that surely looked like the students were hullucinating when they wrote it :-)

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From Lawrence D'Oliveiro@21:1/5 to Jim Jackson on Tue Apr 29 02:41:37 2025
    On Mon, 28 Apr 2025 12:42:26 -0000 (UTC), Jim Jackson wrote:

    Mind you, way back in the day I saw degree student's software that
    surely looked like the students were hullucinating when they wrote it
    :-)

    I wonder how that can be, assuming the code actually works.

    I knew someone, back in undergrad days, who was a good programmer (a few
    years senior to me), but a lousy speller. I remember seeing one of his
    programs reporting that the user had typed an “illegial charector” ...

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From Jim Jackson@21:1/5 to Lawrence D'Oliveiro on Tue Apr 29 13:10:31 2025
    On 2025-04-29, Lawrence D'Oliveiro <ldo@nz.invalid> wrote:
    On Mon, 28 Apr 2025 12:42:26 -0000 (UTC), Jim Jackson wrote:

    Mind you, way back in the day I saw degree student's software that
    surely looked like the students were hullucinating when they wrote it
    :-)

    I wonder how that can be, assuming the code actually works.

    Who said anything about working?

    I knew someone, back in undergrad days, who was a good programmer (a few years senior to me), but a lousy speller. I remember seeing one of his programs reporting that the user had typed an ???illegial charector??? ...

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From Jim Jackson@21:1/5 to Jim Jackson on Tue Apr 29 13:23:16 2025
    On 2025-04-29, Jim Jackson <jj@franjam.org.uk> wrote:
    On 2025-04-29, Lawrence D'Oliveiro <ldo@nz.invalid> wrote:
    On Mon, 28 Apr 2025 12:42:26 -0000 (UTC), Jim Jackson wrote:

    Mind you, way back in the day I saw degree student's software that
    surely looked like the students were hullucinating when they wrote it
    :-)

    I wonder how that can be, assuming the code actually works.

    Who said anything about working?


    Actually it probably did for at least one path through the code! The one
    they actually tested. But this one path went round the houses to get
    there. :-)

    There is a technique to programming where, when it fails to do the job,
    the "coder" (and I use the term loosely) does an edit or add "one
    thing", recompile, test, rinse repeat cycle. Never delete anything, it
    might be important. Code review - don't yer just love it.

    Do AI bots ever compile or run their code? Silly question.
    Could we get another AI to do code review? Actually not a silly question.


    I knew someone, back in undergrad days, who was a good programmer (a few
    years senior to me), but a lousy speller. I remember seeing one of his
    programs reporting that the user had typed an ???illegial charector??? ...

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)