In the span of 24 hours this month, OpenAI and DeepSeek made opposite bets on what frontier AI is worth.

One bet, by OpenAI, was that the future would be an expensive, closed product. The other wager, by DeepSeek, is an open infrastructure that will be cheap. The gap between the two ends of the market is now wider than it has been in years, and the middle class of AI production that most coding agents have been routing through is thinning out, reports TNS contributor Janakiram MSV in today’s lead story.

Until this month, you could pick a model on a fairly smooth price-performance curve. There was a top tier, a middle tier, and a budget tier, and most workloads found a comfortable home somewhere on the slope. That curve still exists, but it has stretched. What used to be a continuous gradient now looks more like two clusters with a gap in between, in the shape of a barbell.

Meanwhile, developers building agents, coding assistants, and high-volume inference pipelines now have to think harder about which side to route to.

Go deeper: The disappearing AI middle class

More interesting reads:

◇ GitHub veteran Brian Douglas launches Paper Compute to fix AI agent infrastructure

◇ The one Slack message that proved our elite engineering team was flying blind

◇ The real story from OpenAI’s big week is Workspace Agents, not GPT-5.5

For years, rules-based automation has only solved 20% of IT problems. The remaining 80% is still costing the industry $250 billion in manual “stitching” and human toil just to keep the lights on. Join us live tomorrow, April 28, to see how shifting from reactive to proactive operations can save millions in labor costs and finally solve the “access” problem in modern delivery.

Stop wasting money on overprovisioned clusters. StormForge uses ML to rightsize pods automatically, matching resources to actual demand while improving performance. Put your optimization on autopilot.