AI adjacent SuperHub.com declines DeepSeek pursuit
illi News/10291190

CHICAGO - illiNews -- For 28 years the end goal for SuperHub.com has always been one of global inclusion. Conceptually this has always included a large silo that housed an efficient, cross-border fusion of Retail and Social. This focus goes back decades to before Google existed.

But there's a new monster in town and they were interested in the hub.

Project Manager Tal Newhart reports, "We were approached by a large Asian developer with a tarnished brand. Their comprehensive strategy was to use outward facing SuperHub.com to combine a developing AI effort, along with widely established retail and social platforms, which they control, to further expand globally. The end thinking  was to use the easy to understand simple English of SuperHub.com, a clean/blank brand, to spruce up their optics and use it for wide-area (read global) data capture and likely leading to sponsored surveillance. Functionally SuperHub.com would become a border-less AI training silo using owned data from unsuspecting retail and social users.

More on illi News

• WOA Crypto redefines cloud Bitcoin mining, offering free access and instant profit opportunities
• Pepperdine University Malibu, California and Community Partners Recognized with 2025 ReadyCommunities Partnership National Service Award
• NBA Overachievers: OddsTrader Reveals Which Teams Will Exceed Expectations in 2025-26
• Crypto Betting Odds 2025: Bookmakers Review Analyzes Market Predictions and Year-End Price Lines
• The OpenSSL Corporation and the OpenSSL Foundation Celebrate the Success of the Inaugural OpenSSL Conference in Prague

Initially the non-trivial offer was enticing. Unfortunately, both the potential buyer's AI, social and retail teams (all smart and experienced) expressed zero sensitivity to the substantial environmental impact of their plans. When asked, they claimed extreme efficiency in building their foundation models adding it was "not likely" that they would ever require hyper-scale data centers adding, ominously, "at least in the near term".

This was a bit like standing in a small room with an elephant sitting in the corner and pretending it's not there.

Newhart adds: "Our ownership can never be considered tree-huggers. That said, we do keep a side-eye on environmental issues with a focus on obvious detrimental impacts. And there is nothing particularly green about some flavors of AI.  Although it's not currently as environmentally hostile as Bitcoin, there is no denying developing large foundation AI models will consume power potentially so massive it will require State support to implement, support this group would certainly be granted. The electrical power, especially in the non-build near term, will need to come from somewhere, and that will have a negative supply and cost impact on local populations. Not to mention water supplies: It takes a huge amount of water to cool the required GPUs and water scarcity is a growing worldwide problem.

More on illi News

• TKL Group's New Factory Commences Production, Pioneering A New Era In Global Heavy Duty Truck Parts
• Regulated Crypto Exchange TZNXG Addresses Core US Market Challenges with Compliance-First Infrastructure
• Unleash Your Inner Seduction at HandPickedSC.com's "The Hot & The Wicked Halloween Soirée"
• GitKraken Launches Insights to Help Engineering Leaders Quantify AI Impact and Improve Developer Experience
• ZapperBox NextGen TV Gateway Receiver Now Testing Support For Secure Whole-Home Content Distribution

Their counterargument is that the likes of TikTok, WeChat and Instagram are all profligate "energy hogs." Perhaps so. But those platforms run nicely on existing infrastructure and don't require expansion of energy grids to feed the hellaciously hungry AI training beast.

Newhart concluded: "Sometimes you just have to take a pass."