We asked how Grok sees VC funding for sensing. (It’s nuanced.)

Ask an AI chatbot a complex question or set of questions and you may get a response that boggles the mind–or at least a human mind. Reporters everywhere are trying to find ways to ethically include AI in research and reporting and writing, and so here’s an interesting example of how I used Grok to explore where VC funding in sensing technology is headed. 

The answer Grok gave, as well as my prompt, are repeated below in full. It came after I wrote a story on the topic with plenty of examples–-all vetted by me—of VCs making investments and of companies receiving the investments as well as where job opportunities are for engineers. 

Why use Grok? I’ve been using Grok more than some others, like Chat GPT or Claude or Gemini. For some reason, I’ve used Grok as a go-to, mainly for getting ideas and art.  Questex, the parent of Fierce Sensors and others, has encouraged reporters and editors to find out more about AI-powered virtual assistants and I mostly want to understand the ways AI has evolved, largely because engineers featured in Fierce Sensors are using AI quite often directly at the edge for physical AI like robots and more. 

(On a less serious note, I have been fascinated at how I came away from reading Elon Musk by Walter Isaacson in 2023 with a rather positive feeling about Musk and his tech insights, including his later work with Grok. Lately, however, I wish Isaacson would do a follow-up expose on how Musk, post-2023, seems to have gone off the rails.)

Before reading the interaction with Grok below, readers should know what I now realize, that the prompt was written quickly and too loosely phrased. It’s like during an interview: Best not to ask two or three questions at once, especially when interviewing a CEO of a tech company! 

At the end of the day, I think the best thing Grok clarified was that “VC funding for sensing-related technologies…is not a massive standalone category like pure AI software of biotech.”  Which makes evaluating the category and measuring it very difficult, even for VC managers and analysts! Grok also noted that discrepancies in dollars invested in sensing tech “are mostly definitional noise in how data providers slice the pie.”  YES! 

If you are still reading and care about AI chatbot nuance, here’s my prompt and Grok’s response.  Here’s also a link to the Fierce story that ran earlier which prompted this entire Grok inquiry.  

PROMPT:  

What is actually happening, Grok, with VC funding for sensing-related technologies? Some of the numbers are showing investments far out of line with others. Does it matter that a company goes public as a measure of its success? Was VC investing really down in 2025 as some report?

GROK: 

VC funding for sensing-related technologies (sensors, sensor fusion, edge AI perception tech like lidar/radar/vision/biosensors, quantum sensing, etc.) is seeing a clear uptick, especially tied to “Physical AI” (robotics, autonomy, and real-world AI applications), defense, and industrial needs.** It’s not a massive standalone category like pure AI software or biotech, but it’s growing steadily as enabling hardware/software for bigger trends. Aggregate numbers for the “sensors” sector hit roughly **$868 million** globally in equity funding across ~53 rounds in 2025 (per Tracxn data through December), up significantly from **$533 million** in 2024 and **$678 million** in 2023.

Notable late-2025 examples include Teradar’s **$150 million** Series B for terahertz vision sensors (automotive/defense/healthcare applications) and smaller rounds like TiHive (~$9M) and Cephia ($4M seed) for multimodal/metasurface sensing. Broader 2025-2026 activity (via VC funds like Lux Capital’s $1.5B raise and Khosla Ventures’ $3B AI-focused fund) includes big checks into robotics/autonomy plays that heavily rely on sensing: Apptronik ($520M for humanoid robots with advanced sensing/reactivity), Wayve ($1.2B autonomous driving AI), Waabi ($70M AI trucking), Aeva (4D lidar), Anduril (sensor fusion in defense AI platforms), and biosensors like Xsensio ($7M).

Industry forecasts point to continued growth: the global sensor market (revenue, not just VC) is projected to exceed **$280 billion** in 2026 in some estimates (up ~8% from 2025), with IDTechEx at a more conservative path to **$250 billion+ by 2036**. VC insiders describe 2026 as “on an upward path,” driven by Physical AI’s need for better motion/location/radar/lidar/optical sensors, sensor fusion, edge AI processing, lower-power always-on designs, and emerging tech like quantum sensing or silicon photonics.

**Why the wild discrepancies in the numbers you’re seeing?** “Sensing-related” isn’t a tightly standardized VC bucket like “AI” or “fintech.” Different reports (Tracxn, PitchBook, Crunchbase, CB Insights, etc.) classify differently:

— Pure hardware sensors vs. software-defined sensing, AI perception layers, or sensor fusion.

– Narrow (e.g., only MEMS/image sensors) vs. broad (embedded in IoT, automotive ADAS, medtech biosensors, defense, climate/environmental sensors, or robotics).

— Early-stage only vs. all stages (including growth equity).

— US-only vs. global, or including/excluding corporate VC/defense tech.

— Hardware risks (inventory, scalability) mean VCs often invest via broader AI/robotics/defense theses rather than “sensors” alone, so a $500M+ robotics round might include heavy sensing components but get bucketed elsewhere.

This leads to apples-to-oranges comparisons—e.g., one source might show modest hardware-only totals while another folds in AI-adjacent perception tech. The sector is also fragmented across verticals (auto, health, industrial, defense), so aggregates vary by data provider and inclusion rules. No single report captures everything perfectly, which explains the “far out of line” figures.

**Overall VC investing in 2025 was *not* down—it surged.** Global venture/growth funding reached **$425–469 billion** (third-highest year on record, up 30–47% from $328B in 2024), per Crunchbase and CB Insights. AI dominated (often 40–70%+ of total, with $211B+ into AI companies), creating mega-round concentration and pushing totals higher even as deal counts fell. US share grew to ~64%. However:

— Early/seed-stage and non-AI deals were softer.

— **VC fundraising** (LPs committing new capital to VC *funds*, not deployments to startups) *was* down sharply—e.g., US VC funds raised ~35% less, near decade lows.

— Some reports focus on this or on Q2/Q3 dips before year-end strength, or exclude the biggest AI outliers. That’s likely what “some report” meant by “down”—context matters (deployed capital vs. new fund commitments; total vs. ex-AI).

Sensing tech benefited indirectly from the AI boom (as Physical AI/robotics/autonomy need real-world sensing) and defense/climate tailwinds, even if it’s a smaller slice overall.

**Does going public matter as a measure of success?** Yes, but it’s not the *only* or even the most common one—especially for VC returns. IPOs are a premier exit: they deliver liquidity for investors/employees/founders, credibility/validation, and access to public capital for scaling. VC-backed IPOs have historically shown strong long-term performance (positive alphas when VCs stay involved for monitoring/governance), and they often generate the biggest multiples in hot markets. In 2025, exits were robust overall (~$549B globally), with public listings accounting for about half the exit value.

That said:

— Most VC exits are still via M&A (acquisitions), which can be faster and cleaner.

— In AI/deep-tech waves, companies often stay private longer (higher private valuations, less regulatory pressure, continued VC support) and still deliver huge returns.

— Success for VCs is ultimately measured by **returns** (IRR, DPI/TVPI multiples on capital), not exit type. A strong acquisition can outperform a mediocre IPO. Going public *does* signal maturity and de-risking for LPs, but prolonged private funding (enabled by abundant VC capital) has let many “unicorns” scale without it.

**Bottom line** Sensing tech VC is healthy and accelerating into 2026 thanks to AI’s shift from chatbots to physical/world-interacting systems—exactly where diverse, accurate, edge-processed sensing becomes critical. The broader 2025 VC market roared back (AI-fueled), contrary to any “down” narrative on deployments (though fundraising lagged). Discrepancies are mostly definitional noise in how data providers slice the pie. IPOs remain a gold-standard success marker for liquidity and prestige, but flexible private exits keep the ecosystem humming. If you share the specific conflicting numbers/reports you saw, I can dig deeper into the exact classification differences!