Loud demos, quiet reality

On specific motorway corridors, in permissive jurisdictions, yes — driver-out freight runs are happening commercially. Everywhere else, and for the messy first and last miles around depots and city streets, a human is still in the cab.

Yes, if you are extremely wealthy. Suborbital hops and private orbital missions genuinely fly paying customers. It is not tourism in any normal sense: seats are scarce, six to eight figures, and manifests are booked years out.

Barely, and this is becoming the important question. As agents log into services and spend money, sites need to know which agent is acting, for whom, and within what mandate. The standards exist in draft; the deployed reality is mostly API keys and hope.

Barely. A small number of countries have approved cultivated meat products, sold in limited restaurants or trial retail. Production volumes are tiny and costs remain far above conventional meat, so this is a novelty rather than a supply chain.

In research, users genuinely feel touch through nerve-interfaced prosthetics, and it measurably improves control. Commercially available advanced prosthetics offer excellent grip patterns but only limited or no true sensory feedback, and cost is a major barrier.

In pilots, under supervision, on narrow tasks. Several manufacturers have humanoids moving totes and feeding lines in real facilities, which is a genuine step up from stage demos. Throughput, uptime and cost per pick are still nowhere near the wheeled robots and fixed automation already doing this work.

For contrived benchmarks, sometimes; for work you would actually pay to have done, not yet. Error rates remain the wall. Error correction has improved enough that the roadmap looks credible rather than mystical, but commercially valuable quantum advantage has not arrived.

In a lab, yes. In your house, no. Research and demo robots fold towels and shirts at slow speed under good lighting, and general-purpose humanoids show it off constantly. Nothing you can buy will reliably empty a real basket of tangled mixed laundry.

For a small number of clinical trial participants, yes — and it is life-changing. Implanted interfaces let people with paralysis move cursors, type and control devices by intent. For everyone else, non-invasive consumer headsets remain crude: a few commands, lots of calibration.

Not as a normal service. Electric air taxis have flown crewed test flights and secured early certification steps, with launch routes announced in a few cities. Commercial passenger operations at meaningful scale have not started.

No. Single-purpose home robots — vacuums, mowers, window cleaners — are genuinely good. A general machine that picks up whatever your family left on the floor, decides where it belongs and puts it away is a research problem, not a product.

No. Humanoids are advancing quickly in warehouses and pilot deployments, but general household work is unsolved. Early units that exist are expensive, supervised, narrow in skill and mostly sold to businesses or developers.

No. Researchers print tissue patches, cartilage, skin grafts and organoids used for drug testing, and a few simple structures have been implanted in trials. A printed, vascularised, transplantable kidney or heart is not close.

No. Current machines are far too small and noisy to factor real keys. The genuine near-term risk is 'harvest now, decrypt later', which is why post-quantum cryptography migration is already underway in serious organisations.

No. Experiments have achieved scientific energy gain in the reaction itself, which is a genuine milestone, but no facility produces net electricity for a grid. Commercial pilot plants are announced for the 2030s and those timelines have slipped before.