Deutsch   English   Français   Italiano  
<v47q52$kk2i$1@dont-email.me>

View for Bookmarking (what is this?)
Look up another Usenet article

Path: ...!eternal-september.org!feeder3.eternal-september.org!news.eternal-september.org!.POSTED!not-for-mail
From: Alan Browne <singularity@blackhole.org>
Newsgroups: misc.phone.mobile.iphone,comp.sys.mac.system
Subject: Re: AI: "Apple Intelligence" - personal and private + ChatGPT 4o
Date: Mon, 10 Jun 2024 17:10:58 -0400
Organization: A noiseless patient Spider
Lines: 113
Message-ID: <v47q52$kk2i$1@dont-email.me>
References: <v47hir$idv0$3@dont-email.me> <v47p7q$khnh$1@dont-email.me>
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8; format=flowed
Content-Transfer-Encoding: 8bit
Injection-Date: Mon, 10 Jun 2024 23:10:58 +0200 (CEST)
Injection-Info: dont-email.me; posting-host="adbbf84814a269f26f569419a762e839";
	logging-data="675922"; mail-complaints-to="abuse@eternal-september.org";	posting-account="U2FsdGVkX1/g+aTX1+hX4kvXcG+hPyFapPWpsFMCijU="
User-Agent: Mozilla Thunderbird
Cancel-Lock: sha1:w928/pgLh0D0ju744vevopN8AgA=
In-Reply-To: <v47p7q$khnh$1@dont-email.me>
Content-Language: en-US
Bytes: 6384

On 2024-06-10 16:55, Your Name wrote:
> On 2024-06-10 18:44:43 +0000, Alan Browne said:
> 
>> Apple announce their AI as "Apple Intelligence" with the emphasis on 
>> _personal_ and _private_.
>>
>> Integrated of course across Apple apps.  (... and 3rd party via App 
>> Intents API).
>>
>> Strong "action" oriented commands.
>>
>> AI will be on device for most processes.
>>
>> Private Cloud Compute for loads that can't be done on device - using 
>> Apple Silicon based servers.  No storage of your data.
>> Independent 3rd party source code verification.
>>
>> Siri to get AI - and boy does she need it.  (Recent use has come up 
>> with a lot of flaws).  Also remembers the recent context in Siri as 
>> well as your cross app context over time.
>>
>> Writing tools for various apps to rewrite messages (such as a mail 
>> message) in different ways.  Grammar checking, tone, etc.  Auto TL;DR 
>> summary generator, etc. and so on.
>>
>> Reading tools to summarize incoming e-mails with relevant content.
>>
>> Emoji generator (text what you want, it gens an "emoji" like 
>> illustration.  (silly fun).
>>
>> ChatGPT 4o integration w/o an account.  If you have a ChatGPT account, 
>> then "paywall" features will be accessible as well...
>>
>> (As soon as the Keynote mentioned ChatGPT - Apple stock shot up).
>>
>> Looking forward it.  Need devices compatible with the latest OS' of 
>> course.
>>
>> Ref: WWDC - 2024 - Keynote (YouTube).
> 
> And *nobody* even knows why or what they would want the useless crap 
> actually for ... the geek brigade just want it so big tech is happy to 
> add it to everything and charge you more for doing it.  :-\
> 
> Yet more pointless gimmick "features" added while still not fixing all 
> the annoying bugs.

I recently read the Ken Follett book, "The Armour of Light" and it 
mentioned the Luddite uprising against industrialization.

You were first to come to mind!


I've been using ChatGPT for a couple years, it's very good.  An example: 
I wrote a long letter in French but found it too casual for the intended 
audience.  I had ChatGPT change it to be more serious in tone.  It 
employed French formality a little too strongly, so I asked it to be 
more Québecois in tone (and still formal).  It handled it beautifully.

Here's another nice example:
"Please compose a three paragraph essay comparing large language models 
to the admonishment in "Dune" forbidding the making of machines in the 
likeness of the human mind."

Result:
"Large language models, such as GPT-3, share an intriguing parallel with 
the cautionary tale presented in Frank Herbert's "Dune" regarding the 
prohibition against creating machines in the likeness of the human mind. 
In "Dune," the Butlerian Jihad has left a lasting imprint on the 
universe, warning against the dangers of relying too heavily on 
artificial intelligence. Similarly, large language models have raised 
concerns about their potential consequences. While not machines in the 
traditional sense, these models, like GPT-3, possess an immense capacity 
to understand and generate human-like text. The cautionary undertone in 
"Dune" echoes in contemporary discussions about the ethical implications 
and control over powerful language models, hinting at the potential 
risks of unbridled reliance on advanced artificial intelligence.

Both the Butlerian Jihad in "Dune" and the debates surrounding large 
language models grapple with the idea of preserving human agency and 
preventing the undue concentration of power. In Herbert's universe, the 
ban on thinking machines aims to safeguard human independence and 
prevent the creation of entities that might surpass or subjugate 
humanity. In a similar vein, concerns surrounding large language models 
center on the concentration of knowledge, influence, and decision-making 
capabilities within the hands of a few organizations or individuals. The 
comparison prompts reflection on the delicate balance between 
technological progress and the need to ensure that such advancements 
align with human values and autonomy.

Despite the cautionary elements, large language models like GPT-3 also 
offer significant potential for positive impact. They have the capacity 
to enhance communication, generate creative content, and aid in various 
tasks. However, the key lies in developing responsible frameworks and 
ethical guidelines to govern their use. Just as the universe in "Dune" 
navigates the aftermath of the Butlerian Jihad, our society must grapple 
with the challenges posed by advanced language models and work towards 
harnessing their potential while ensuring safeguards against unintended 
consequences. Ultimately, the comparison invites reflection on the 
responsibility that comes with wielding powerful technologies and the 
imperative to shape their trajectory in harmony with human values."


Which ain't bad.  College profs have to use software to detect when 
their students are handing (e-mailing in) assignments to be sure they're 
not AI generated...


-- 
"It would be a measureless disaster if Russian barbarism overlaid
  the culture and independence of the ancient States of Europe."
Winston Churchill