
Microsoft has turned yield into the key test for AI.
Yield, a semiconductor industry measure, is now applied to AI.
AI will be judged by yield, Microsoft says.
What happened
Microsoft on Wednesday applied the semiconductor industry's oldest measure of success—yield—to AI, arguing that AI will be judged by it.
Why it matters
Yield, long the semiconductor industry's benchmark, is now being positioned by Microsoft as the standard for AI, signaling a shift in how AI's value is assessed.
What to watch
Watch for how the industry responds to Microsoft's framing, as yield becomes a potential new metric for AI's practical success.
Ask the AI about this article →
Summaries like this, in your inbox every morning.
Microsoft's move to apply yield—a concept long tied to semiconductor manufacturing—to AI marks a notable rhetorical shift. Yield, traditionally a measure of production efficiency, is now being repurposed as a benchmark for AI's success. This could signal a broader industry trend toward evaluating AI not just on capabilities but on more practical, measurable outcomes. However, the article provides no further details or reactions, leaving the full implications open to interpretation.
For example, today's edition would include:
AI-summarized, only the topics you pick — one digest a day via Email, Slack, or Discord.
Free · 30 seconds with Google · unsubscribe anytimeWhat is AIToday? →
Ask AI anything about this article. Q&As are published on this page for other readers too.
Tesla Japan will show the Cybercab, its dedicated Robotaxi vehicle, in Japan for the first time

Palantir Technologies Inc

Researchers in Bengaluru, India, are using AI to train rescue dogs like Chloe to sniff human breath for cancer…

JM-Applied, a Taiwanese semiconductor gas equipment maker and supply chain member for Micron and TSMC, said on…

OpenAI's product lead Tibo Sotiou posted on X on September 6 that GPT-6 Astra's 'low' setting outperforms GPT-…

Nvidia disclosed roughly $99 billion of public and private equity investments as of July 26, plus about $25 bi…
