As of late 2025, the current pattern of AI‑enabled robot deployment is strongly consistent with Sacks’ prediction, but his stated horizon (“in the near term” / “over the next several years, into the late 2020s”) has not finished yet, so it’s too early to declare it definitively right or wrong.

What we see so far (2023–2025):

• Industrial robots massively dominate robot deployment: the International Federation of Robotics (IFR) reports a record 4.28 million robots operating on factory floors worldwide in 2023, with more than 540,000 new industrial units installed annually in 2023–2024.(linkedin.com) This is orders of magnitude larger than any other robot category.
• Professional service robots are growing, but they’re also mostly industrial or commercial, not domestic: in 2023 over 205,000 professional service robots were sold, and more than half were for transportation and logistics (warehouse/industrial handling), with other big segments in hospitality, cleaning, agriculture, and security.(ifr.org) These are deployed in workplaces, malls, hospitals, airports, etc., not as general‑purpose home helpers.
• Humanoid and “advanced” robots in 2024–2025 are primarily piloted in factories and logistics. An IDTechEx market study notes that, as of 2025, humanoid robots in the automotive sector are still in early pilot tests, doing basic material‑handling and inspection, and that general‑purpose humanoids in non‑industrial areas (like healthcare) are even further away; warehouse pilots number fewer than 100 humanoid units.(idtechex.com) A Wired piece similarly frames 2025 as the year of the humanoid robot factory worker, with Boston Dynamics’ Atlas going into a Hyundai plant.(wired.com) A deployment survey of humanoids shows they are working mainly in logistics, security, elder care, retail, and R&D pilots—again, institutional rather than home use.(mechonomics.co)
• Military use of robots and drones is substantial and growing: Ukraine is creating dedicated units of robotic ground vehicles; Russia has formed a large drone‑warfare unit (“Rubicon”); and the US and allies are developing collaborative combat drones like the XQ‑58 Valkyrie and long‑endurance UAVs such as ULTRA.(reuters.com) This supports the prediction’s mention of military applications.
• By contrast, general‑purpose robots in private homes are still rare and experimental. A notable example, 1X’s NEO “home robot,” only opened pre‑orders in October 2025, targets early adopters at about $20,000 (or $499/month), and currently relies heavily on remote human teleoperation rather than full autonomy—indicating an early, niche market rather than broad deployment.(en.wikipedia.org) Analysts also project that truly general‑purpose humanoid robots for non‑industrial domains will not begin to scale until later in the decade.(idtechex.com)

Why the verdict is “inconclusive”:

• Sacks’ directional claim—that in the near term most practical deployment of advanced AI robots would be industrial or military rather than as general‑purpose home robots—matches the evidence up to November 2025.
• However, the normalized prediction explicitly extends “through at least the late 2020s.” We have not yet reached that point, and rapid breakthroughs or unexpected consumer adoption (e.g., affordable, capable home humanoids post‑2026) could still change the balance.

So, while the world in 2024–2025 looks very much like what he described, the multi‑year horizon is still unfolding, making the overall prediction inconclusive (too early to fully score).