Sarah and all
just one more quick sunday morning reflection from me, following my last
post

Based on what I am  experiencing
humans must figure out how to learn faster than machines to maintain some
level of control and understanding over what is going on
I am finding it quite challenging, while trying to hang onto my own mental
faculties

Great weekend all
 :-)


On Sat, Mar 29, 2025 at 11:38 AM Paola Di Maio <[email protected]>
wrote:

> Sarah and everyone
>
> I apologise for generalizing and possibly trivializing the discussion
>
> What I think I am trying to say is that:
> - experts like you and other Carpenters are scarce, the majority of
> learners may not be able to write good code as efficiently as you do
> nor to be able to tell what is good code,
> - an analogy with essay writing: students are using AI to write their
> essay, thus possibly losing their ability to compose text without AI
> -Since the printing press became a thing, things have been moving fast -
> but of course we are still learning how to write by hand using pen and
> paper, as this is part of our cognitive development, How long do you think
> children will be taught how to write by hand in schools using pen and
> paper? *i wonder
>
> - AI is learning FAST very FAST. I am not a technocrat, but I am reaping
> the benefits of this advancement. I embrace what I can of it
> and try to metabolize it critically. Let s not forget that these ai code
> generators are just first generation. I expect rapid advances
>
> - AI can be useful if we can learn from it *can we learn to write
> good/better code by using as a starting point the poorly formed output of a
> code generator? can we learn about code generators as well by doing so?
>
> - working with AI may require metacognitive abilities *in coding, in essay
> writing etc which may be even more important
> to develop acquire than coding skills in themselves
>
> - given the above humans need to continue to develop and use their
> intelligent faculties and learned skill to keep on top of machine
>
> It would be awesome to see  examples or even a book/manual from the
> excellent instructors on this list
> something like:
> Given a set of prompts in AI code generator
> get ai generated code
> Hold Nose
> analyze it and evaluate it critically
> write it better yourself
>
> Look forward to continued learning
> Thank you
> P
>
>
> -
>
> On Fri, Mar 28, 2025 at 9:30 PM Sarah Brown <[email protected]> wrote:
>
>> Thanks Paola for your perspective on the goal of the conversation.
>> Understanding that you had a different scope makes your comments make more
>> sense to me.  Since the original message pointed us to the PR, I still feel
>> that the current message to go into the curriculum is the top priority.
>>
>> A problem I am aware of is that because these things move so fast, most
>> accessible, nontechnical material that people rely on is corporate
>> marketing materials, which present a very skewed view.
>> My examples come from:
>>
>>    -  (mostly) scholarly literature in communities like CHI (top venue
>>    in human computer interaction) and FAccT (a top, albeit new, venue on
>>    fairness, Accountability and transparency in computing) among others and 
>> my
>>    department's recent talks by CS PhD candidates and post docs during a 
>> round
>>    of hiring
>>    - (less) from my second hand experience of seeing undergraduate
>>    computer science students' learning undermined by them using LLMs when 
>> they
>>    are novices and then they are completely unable to learn material in later
>>    courses. Even in supervising research students, students' reliance on LLMs
>>    has made my work of code review much more exhausting because LLMs code
>>    style is not great in general and students on their own gave me code I was
>>    much happier with than what they get with LLM help.
>>    - (minimal) first hand experience using and (more) stress testing
>>    LLMs. I have almost compeltely stopped trying to use them for my own
>>    programming work because after a few tries they gave too much code of such
>>    low quality that they slowed me down, I can get my work done faster 
>> without
>>    the LLM in most cases. The exception is when I work in a language I do not
>>    know well, they provide the kind of starting point that I used to get from
>>    stack overflow a little faster than digging through sites like that.
>>
>> I do not have a link to a specific recording, but Ruha Benjamin, at
>> princeton, has been giving a talk lately that groups both utopian and
>> dystopian views of an AI future as technodeterminism, assuming that the
>> future of how a technology impacts society is predetermined, and calls for
>> us to challenge that-- to center people and our collective agency to shape
>> the world we want to live in, rather than assuming that our future world
>> will be handed to us by tech firms. That is what is most important to me. I
>> think the Carpentries community is well positioned to empower people
>> to take this leap, but I am not sure what it looks like to get there.  I
>> hope we can keep having human-first conversations in this community.
>>
>> *Sarah M Brown, PhD*
>> sarahmbrown.org
>> Assistant Professor of Computer Science
>> University of Rhode Island
>>
>>
>>
>> On Sun, Mar 23, 2025 at 6:42 PM Paul Harrison via discuss <
>> [email protected]> wrote:
>>
>>> On Wednesday, March 19, 2025, at 8:38 PM, Toby Hodges wrote:
>>>
>>> If we want to cover this, it needs to be in a separate lesson or as an
>>> almost total rewrite of existing materials IMO. Delving into this in detail
>>> would be too time consuming during a workshop otherwise, at the cost of all
>>> the other important things we want to teach people.
>>>
>>>
>>> Hi Toby. That sounds reasonable, I totally agree, what I was talking
>>> about is beyond the scope of updates to existing lesson material. I also
>>> agree with Sarah Brown's comments.
>>>
>>> This list is for the purpose of general discussion about The Carpentries
>> including community activities, upcoming events, and announcements. Some
>> other lists you may also be interested in include discuss-hpc, discuss-r,
>> and our local groups. Visit https://carpentries.topicbox.com/groups/ to
>> learn more. All activity on this and other Carpentries spaces should abide
>> by The Carpentries Code of Conduct found here:
>> https://docs.carpentries.org/topic_folders/policies/code-of-conduct.html
>> *The Carpentries <https://carpentries.topicbox.com/latest>* / discuss /
>> see discussions <https://carpentries.topicbox.com/groups/discuss> +
>> participants <https://carpentries.topicbox.com/groups/discuss/members> +
>> delivery options
>> <https://carpentries.topicbox.com/groups/discuss/subscription> Permalink
>> <https://carpentries.topicbox.com/groups/discuss/Tc7ad8d3684017cd0-Mf92d81bef41009e956dfd942>
>>

------------------------------------------
This list is for the purpose of general discussion about The Carpentries 
including community activities, upcoming events, and announcements.  Some other 
lists you may also be interested in include discuss-hpc, discuss-r, and  our 
local groups. Visit https://carpentries.topicbox.com/groups/ to learn more. All 
activity on this and other Carpentries spaces should abide by The Carpentries 
Code of Conduct found here: 
https://docs.carpentries.org/topic_folders/policies/code-of-conduct.html

The Carpentries: discuss
Permalink: 
https://carpentries.topicbox.com/groups/discuss/Tc7ad8d3684017cd0-M425969677b122a2cb28d5dbf
Delivery options: https://carpentries.topicbox.com/groups/discuss/subscription

Reply via email to