Technology Evening Edition

Tech Sector: Mixed AI Signals - Jan 25

A weekend of mixed signals in tech: Brex's cut-price exit, new AI model startups, content provenance questions for ChatGPT, and rising data-center power costs. Heading into Monday, investors should weigh innovation against valuation and infrastructure risks.

Sunday, January 25, 20266 min readBy StockAlpha.ai Editorial Team
Tech Sector: Mixed AI Signals - Jan 25

Share this article

Spread the word on social media

The Big Picture

Heading into the long weekend, the technology sector posted a string of contrasting headlines that leave investors with both opportunities and questions. The most impactful development was the Capital One acquisition of Brex for $5.15 billion, a result that is prompting fresh scrutiny of startup valuations and investor expectations.

At the same time, innovation keeps coming, from new AI companies focused on coordination to platforms paying tens of thousands of experts to improve models, even as cultural pushback and infrastructure costs create fresh headwinds. Markets were closed Sunday; the last trading day was Friday, January 23, and the next session opens Monday, January 26.

Market Highlights

Quick facts and the headlines you need to know as you prepare for Monday.

  • Brex, once valued at about $12 billion, will be acquired by Capital One for $5.15 billion, a price roughly 57% below the prior peak valuation, raising questions about funding expectations for AI-era startups. The buyer is $COF.
  • OpenAI's ChatGPT is now returning material that appears to be sourced from Elon Musk's Grokipedia, highlighting content-provenance issues for large language models and potential moderation concerns.
  • Humans&, a startup founded by alumni from Anthropic, Meta, OpenAI, xAI, and DeepMind, says it's building foundation models focused on coordination rather than chat, signaling a fresh product vector in AI research and commercialization.
  • Mercor reportedly pays roughly $2 million per day to about 30,000 experts training AI models, averaging $95 per hour and with some specialists, like radiologists, earning up to $375 per hour, underscoring rising human-in-the-loop costs.
  • Power prices spiked in Virginia, home to the world’s largest data-center hub, with record winter demand expected amid a storm, a near-term cost pressure for cloud and infrastructure operators such as $AMZN, $GOOGL, and $META.
  • Consumer tech coverage included a positive review of a solid-state portable battery and deep discounts on TVs at retailers like Best Buy, where deal-focused traffic may support short-term consumer spending on electronics, $BBY.

Key Developments

Brex sale and the valuation reset

Capital One's $5.15 billion acquisition of Brex is being framed as a top-tier exit, yet it's a marked step down from Brex's roughly $12 billion private valuation. For founders and investors in AI and fintech, this is a reminder that fundraising expectations tied to 'absolute domination' are risky, and you may want to reconsider how much premium the market will pay for growth claims alone.

AI sourcing, new models, and cultural resistance

ChatGPT pulling answers from Grokipedia puts the spotlight on training data provenance and ideological influence in model outputs. At the same time, startups such as Humans& are pushing into coordination-focused models, and Mercor's large-scale expert pay shows demand for human oversight. But cultural backlash continues, with some creators and events taking firm stands against generative AI, which could influence public perception and licensing access for creative content.

Infrastructure strain and the economics of model training

Rising power prices in Virginia during a winter storm show how physical infrastructure can become a short-term constraint on cloud operations. Combine that with the daily multimillion-dollar payouts Mercor is making to experts and you get growing operational costs for training and running advanced models. That could squeeze margins for both startups and incumbent cloud providers unless companies pass costs on or find efficiencies.

What to Watch

As markets reopen Monday, you'll want to track a few clear catalysts and risks that will shape tech sector performance in the near term.

  • Monday market reaction, earnings, and guidance: watch how investors price valuation-related stories like Brex and any reported guidance changes from cloud and fintech firms. The next trading day is Monday, January 26.
  • Data and content provenance: monitor follow-up reporting or statements from OpenAI, xAI, and other model providers about data sources and moderation. Could regulators or partners demand clearer attribution?
  • Infrastructure costs: keep an eye on regional power prices and data-center uptime reports, especially in Virginia. How will $AMZN, $GOOGL, and $META manage energy-driven costs?
  • Labor cost dynamics: track Mercor and similar services to see whether high pay for experts becomes a sustainable model or a sign of rising marginal costs for model improvement.
  • Policy and reputation: will events and creator pushback lead to new licensing demands or voluntary restrictions on AI training data? That could affect content pipelines and product roadmaps.

Bottom Line

  • Valuations are being reassessed, as the Brex sale shows, so manage position sizes and expectations if you own high-flying private-to-public transition plays.
  • Innovation remains strong, with new AI startups and paid expert networks suggesting ongoing product and capability advances, but these come with rising costs.
  • Operational risks matter now, with energy and expert labor costs likely to affect margins for cloud and AI businesses in the near term.
  • Content provenance and cultural pushback are material risks for model providers and platforms, so watch regulatory and partner responses.
  • Be selective and stay nimble, because the sector is offering both long-term upside and short-term volatility as narratives evolve.

FAQ Section

Q: What does the Brex sale mean for startup valuations? A: It signals a reset for some late-stage private valuations and reminds investors that lofty multiples tied to market dominance can be repriced if strategic exits come at lower levels.

Q: Should you be worried that ChatGPT is using Grokipedia content? A: You should monitor responses and vendor statements, because data provenance affects reliability and potential moderation or licensing issues, but one data source alone does not invalidate a model.

Q: How could higher power and expert-pay costs affect tech companies? A: Higher operational costs can pressure margins, especially for cloud and AI-heavy businesses, so watch near-term guidance and capital allocation decisions from major providers.

Sources (10)

#

Related Topics

AIstartup valuationsdata centersMercorpower pricessolid-state batteries

Disclaimer: StockAlpha.ai content is for informational and educational purposes only. It is not personalized investment advice. Sentiment ratings and market analysis reflect data-driven observations, not buy, sell, or hold recommendations. Always consult a qualified financial advisor before making investment decisions. Past performance does not guarantee future results.

Spotted something wrong? Report an error.