In this week's issue:

  • Terafab's power requirements are the actual bottleneck—and they're cracked, not catastrophic.
  • Prediction markets face bipartisan siege, but the real story is Don Jr.'s naked conflicts of interest.
  • TOMO Mags: Austin analog collectors reject the algorithm; print shops are building community infrastructure.
  • AI automation is bifurcating labor: experienced workers rise, entry-level workers disappear.

Let's go.

Top Stories

Terafab's Power Problem Isn't Actually a Problem

The Terafab story's dirty secret is that the 10+ gigawatt power requirement—often framed as a dealbreaker—is solvable through mechanisms already deployed across Texas's data center boom. Here's the operational reality beneath the hype.

Terafab targets 1 terawatt of annual compute output. At full scale, peak loads exceed 10 GW. ERCOT's current peak capacity is approximately 70 GW, with a projected 6.2% deficit in summer 2026. A 10+ GW new load increases total demand by ~14%—substantial, but not impossible. The trick: Musk designed Terafab with an 80/20 orbital strategy. Eighty percent of Terafab's chips ship to space-based AI data centers powered by solar and radiators—they never touch Earth's grid. Only 20% of production lands in terrestrial applications (Optimus, vehicles, Earth-based AI). That means Terafab's sustained grid load is closer to 2-3 GW, not 10+ GW, dramatically reducing strain. The remaining compute stays orbital, avoiding grid constraint entirely.

On-site power generation is already standard practice in hyperscale data center infrastructure. Bloom Energy is deploying fuel cell systems across Texas; NextEra Energy is building dedicated 500+ MW natural gas turbine plants for data center clients; Google and Meta are contracting massive solar installations. Texas's SB6 legislation explicitly encourages large-load users to build integrated power solutions by shifting grid interconnection costs and reliability risks to the facility operator. This isn't new: companies like Apple and Facebook have operated their own power plants for a decade. Terafab will do the same—fuel cells, natural gas turbines, battery storage, solar. Multiple redundancy, zero grid exposure. The permitting timeline is real (2027-2028+), but the technology isn't speculative.

The actual bottleneck is land and permitting timelines, not physics. Terafab requires thousands of acres and multiple years of regulatory clearance. ERCOT's interconnection study process for 1+ GW loads runs several months to over a year. But for Musk's stated goal of small-batch production in late 2026, the footprint is much smaller. The 100M-square-foot megafab will likely land elsewhere in Texas—possibly Central, possibly Corpus Christi area where land is cheaper and power solutions are already in-flight. Austin gets the advanced tech fab; the volume production fab goes where grid and land economics work. This is pro-building Texas in action: no artificial scarcity, no regulatory capture. The market finds the answer.

Sources: ERCOT 2025 Annual Report, ERCOT Grid Interconnection Process, Bloom Energy Fuel Cells, NextEra Energy Gas Plants, Terafab Orbital Strategy, Tesla/SpaceX Announcement

Prediction Markets Are Under Siege—And The Conflicts Run Deep

This week, senators introduced the "Prediction Markets Are Gambling Act," a bipartisan bill banning sports betting contracts on CFTC-regulated platforms. It's a niche legislative move, but it exposes something far uglier: how regulatory capture works at the highest levels when you have undisclosed financial interests and administration access.

The backstory: Polymarket and Kalshi users profited substantially from advance knowledge of Trump administration military actions in Iran and Venezuela, triggering scandal and legitimate insider-trading scrutiny. The platforms implemented guardrails this week—no athletes, no politicians, no one with inside information can trade certain contracts. Performative stuff. But here's the real problem: Donald Trump Jr. is simultaneously an investor in Polymarket (through his VC firm 1789 Capital, $10-20M stake) and a strategic advisor to Kalshi, with undisclosed compensation. He has direct financial upside if the Trump administration's CFTC—which backs these platforms—continues sheltering them from state-level gambling prosecutions and federal restrictions.

The Trump administration already dropped two Biden-era federal investigations into Polymarket. Don Jr. profits if that leniency continues. The bill from Schiff and Curtis is early-stage and narrowly focused (sports betting only), but it signals congressional appetite to restrict prediction markets more broadly. The platforms are not equally exposed: Kalshi faces criminal gambling charges in Arizona and derives 90% of volume from sports betting—the bill directly targets its revenue model. Polymarket, backed by NYSE parent company, has more institutional protection and diversified markets. But both benefit from Don Jr.'s access and lack of recusal. That's textbook conflict of interest.

The real issue isn't whether prediction markets are good (they're excellent epistemic tools—our newsletter agrees). The issue is whether a formal regulatory review of these platforms can be trusted when the administration's closest advisor to the incoming tech leadership has undisclosed financial stakes in their success. If the markets deserve protection, they should be defended on principle, not insider favor. If they're genuinely risky, that risk should be analyzed honestly, not obscured by regulatory capture. Austin's crypto-positive, pro-prediction-market ecosystem depends on this distinction: innovation should win on merit, not connections.

Sources: Kalshi and Polymarket Insider-Trading Ban, Trump Jr. Polymarket Investment, Trump Jr. Kalshi Advisory Role, Trump Admin Dropped Polymarket Investigations, Schiff-Curtis Bill

  • SXSW 2026 continues through March 16. Film, interactive, and music festival with jury-selected projects, industry panels, and the annual collision of tech, film, and music culture in Austin.
  • JD Vance RNC Fundraiser. Vice President headlines high-dollar GOP event in Austin hosted by tech elite, $100K per couple.
  • Austin Tech Meetup — Data Center and AI Infrastructure. Industry panel on Texas's AI compute buildout, featuring ERCOT representatives and hyperscaler executives, March 28.
  • Barton Springs Sunrise Sessions. Weekly community gathering at Austin's iconic spring-fed pool, casual and free, every Saturday and Sunday morning.

One Thing

If this issue helped you separate signal from noise in Austin's tech and political moment, here's how you can help:

  • Forward this newsletter to one Austin builder or creator who should be reading it
  • Reply with your thoughts on Terafab's execution risk or prediction market governance—I read everything
  • Share if you found the grid analysis valuable

Thanks for reading The Austin Daily News. Stay skeptical and build something.

Was this forwarded to you? Subscribe to get it every day.

Keep Reading