Hoi,
You indicate that over a billion people use ChatGPT... the closest analogy
is the Boxer wars where the UK fought to have a market for opium in China.
It was to keep people depending on drugs. AI from a strategic point of view
creates a dependency on the USA. The global narrative is that as a partner,
the USA can no longer be depended on. It is why Europe and China are moving
away from products that are keyed to the US national interest including
products like SWIFT, personal computing, computing.
When some US based megalomanic billionaires outcompete each other to
develop AI, they are a substantial producer of worldwide CO2, this while we
are already past tipping points we should not cross when we want to have a
livable earth. When it is considered a fact that a billion people use
ChatGPT, I wonder if that is because a lot of the hardware and software is
infused with AI.
AI production is not only producing CO2 it is actively using up water where
there is not enough to go around in the quantity desired by AI factories.
There is a reason why people do not want AI factories in their backyard and
there is a reason why these factories are announced when permissions have
already been granted.
Google used to use green energy for its services, it no longer can.. the
same is true for all AI companies.
AI is used in a strategic fashion and it cannot be assumed to be/remain
globally available. We are a global organisation and our public is global.
AI, in its current iteration, is something we cannot afford. When we
provide a search engine with all our references, with inclusion of like
minded organisations, when videos are included and when we enable
annotation to all we offer, we do not offer a substitute for AI but we
offer something substantially more inclusive than just Wikipedia.
Also, when the WMF wants to be green, it can source its energy from a
supplier that provides green energy. When that is not possible, it can
build an equivalent new capacity and provide it on the grid. Then again, in
this argument we consider AI not WMF.
Thanks,
GerardM
On Fri, 10 Jul 2026 at 20:05, Steven Walling <[email protected]>
wrote:
>
> On Fri, Jul 10, 2026 at 10:36 AM Gerard Meijssen via Wikimedia-l <
> [email protected]> wrote:
>
>> Hoi,
>> How does that raise more eyeballs to our data, to our projects. Why would
>> we build this to serve companies that raise CO2 with 35%?
>>
>> Why support AI at all when it is killing our future!
>> Thanks,
>> GerardM
>>
>
> Gerard,
>
> ChatGPT alone has over a billion monthly active users and these platforms
> are growing rapidly (note the survey data I linked to previously in
> thread).
>
> Everything Wikimedia does has a CO2 emissions cost. We are not carbon
> neutral today as far as I am aware, though I would bet many of us would
> agree that is a goal worth pursuing.
>
> Steven
>
>
>> On Fri, 10 Jul 2026 at 19:30, Steven Walling via Wikimedia-l <
>> [email protected]> wrote:
>>
>>>
>>>
>>> On Fri, Jul 10, 2026 at 9:52 AM Erik Moeller via Wikimedia-l <
>>> [email protected]> wrote:
>>>
>>>> On Fri, Jul 10, 2026 at 5:59 PM James Heilman via Wikimedia-l
>>>> <[email protected]> wrote:
>>>>
>>>> > Yah a search engine that actually gives real references that supports
>>>> the statements in question would be amazing.
>>>>
>>>> Almost like .. a Knowledge Engine. ;-)
>>>>
>>>
>>> Is the WMF building an MCP server to connect Wikipedia and Wikidata
>>> directly to Gemini, Claude, and ChatGPT? This is a more lightweight,
>>> backdoor way to leverage the audience of those platforms but present
>>> structured outputs to AI chats for users based on Wikimedia knowledge. If
>>> we did so, we could present citations within the returned responses to
>>> users, and those platforms make it transparent to the user when they are
>>> calling a particular tool.
>>>
>>> Steven Walling
>>>
>>> Sadly, the only realistic path I see there would be through
>>>> acquisition, and even if that was financially feasible, you'd begin by
>>>> inheriting a lot of corporate practices that aren't really consistent
>>>> with Wikimedia values.
>>>>
>>>> But perhaps there is a middle ground where Wikimedia seeks to define
>>>> more clearly the terms of engagement that it wants with search engines
>>>> (clear attribution, clear and correct references, calls-to-edit,
>>>> etc.), and then finds and recognizes search partners who implement
>>>> those. To Luis' point, that need not be done by WMF.
>>>>
>>>> Warmly,
>>>>
>>>> Erik
>>>> _______________________________________________
>>>> Wikimedia-l mailing list -- [email protected],
>>>> guidelines at: https://meta.wikimedia.org/wiki/Mailing_lists/Guidelines
>>>> and https://meta.wikimedia.org/wiki/Wikimedia-l
>>>> Public archives at
>>>> https://lists.wikimedia.org/hyperkitty/list/[email protected]/message/HNDPDZBBDMILF7WGMUBVJVIAZYQ7OXOS/
>>>> To unsubscribe send an email to [email protected]
>>>
>>> _______________________________________________
>>> Wikimedia-l mailing list -- [email protected], guidelines
>>> at: https://meta.wikimedia.org/wiki/Mailing_lists/Guidelines and
>>> https://meta.wikimedia.org/wiki/Wikimedia-l
>>>
>> Public archives at
>>> https://lists.wikimedia.org/hyperkitty/list/[email protected]/message/IUDAKJRN5EQT5CCWEEYQXKDSVXCDBXR6/
>>> To unsubscribe send an email to [email protected]
>>
>> _______________________________________________
>> Wikimedia-l mailing list -- [email protected], guidelines
>> at: https://meta.wikimedia.org/wiki/Mailing_lists/Guidelines and
>> https://meta.wikimedia.org/wiki/Wikimedia-l
>> Public archives at
>> https://lists.wikimedia.org/hyperkitty/list/[email protected]/message/N7LFQLLPT6BSANTYFYPXHHQ6XYQX76GA/
>> To unsubscribe send an email to [email protected]
>
>
_______________________________________________
Wikimedia-l mailing list -- [email protected], guidelines at:
https://meta.wikimedia.org/wiki/Mailing_lists/Guidelines and
https://meta.wikimedia.org/wiki/Wikimedia-l
Public archives at
https://lists.wikimedia.org/hyperkitty/list/[email protected]/message/6IEAISQHBW7ILI34T2M2FNHNBWL76T7Y/
To unsubscribe send an email to [email protected]
