Welcome to
AI runs everything now — but who runs AI?
That’s the question we put to readers last week in our front-page poll: Which battle will define the next decade of tech?
This Week’s Poll Results (HackerNoon)
AI is now the axis of the tech industry. The question is: which force will shape its future the most?
A total of 160 readers weighed in, and the results were.. Surprisingly balanced.
- 31% said it’s all about compute power — whoever controls the chips and data centers calls the shots.
- 28% went for trust & culture, meaning the fight between open-source believers and closed-model empire builders.
- 26% picked consumer entry points — the apps, phones, and chat platforms where AI actually meets humans.
- And 16% went for policy and regulation, the rules that (supposedly) keep the rest in check.
In other words, readers don’t see one winner — they see a standoff.
The top spot went to compute, which makes sense considering whoever owns the GPUs owns the pace of progress. Even nations are thinking in teraflops now: the UK’s “AI Research Resource” project and China’s push for sovereign chip supply show that compute isn’t just a business asset anymore — it’s national strategy.
But the thing about power (literal, electrical power) is that it’s expensive. Training models with trillions of parameters costs more energy than small countries use in a year, and that makes compute a choke point — one that may define who even gets to play in AI by 2030.
The close second-place finish — trust and culture — tells us something important: people are catching on that the real question isn’t just how big your model is, but how you built it.
Meanwhile, 26% of voters picked the battle over consumer entry points. If AI becomes a layer between you and everything else — messages, maps, shopping, work — then the gatekeepers aren’t the GPU makers or the model builders. They’re the ones who own your interface. In that world, the “app” as we know it might disappear. The “assistant” becomes the operating system.
And then there’s policy, which scored lowest but matters most once the champagne wears off. The EU AI Act is finally here, with categories, audits, and risk tiers that’ll make compliance officers the new rockstars.
So what does this all mean? Personally, we think we’re not heading toward a single dominant battle. We’re heading toward interdependence:
- Compute builds capability.
- Culture defines what’s acceptable.
- Entry points decide what’s adopted.
- Rules determine what’s allowed.
They’ll rise and fall in tension — a messy, dynamic balance of power. And that’s exactly how tech progress should look when the stakes are civilization-level, because if one side ever truly wins — if compute monopolies lock everyone else out, or regulators smother innovation, or the open-source ethos dies — AI stops being a creative force and starts being an extractive one.
Speaking of AI, betting on what will be the best AI model appears to be the rage at the moment.
🌐 From Around the Web: Polymarket Pick
Current odds: 50% Google
Last time around, we saw users on Polymarket
Models will be ranked based off of the highest arena score on the
🌐 From Around the Web: Kalshi Pick
Current odds: Anthropic: 68% | Google: 19% | OpenAI: 18%
Meanwhile a cool $100,000 is riding on which AI company will have the best coding model on January 1, 2026, over at
We want to hear from you!
That’s it, folks! We’ll be back next week with more data, more debates, and more donut charts!
In the meantime:
Vote on this week’s poll: What’s Your Biggest Struggle As a Writer in 2025?