Putting a coin-sized chip into the brain and letting someone with severe paralysis move a cursor on a screen by thought alone has already left the laboratory. Elon Musk's brain-computer interface company tells it as a production line that will sooner or later start turning.
Yet a pleasing vision is one thing; a production line has to be delivered step by step, process by process. The cases already on the ground and the plans still on paper sit a fair distance apart, and that gap is what this piece tries to sort out layer by layer.
Moving Again at the Bedside
According to Guangming Online and other media reports, as of September 2025, twelve people with severe paralysis worldwide had received the company's implant. They can steer a cursor with brain signals, type words, play simple games and post on social platforms.
For someone who has lost control of their limbs, that window needs little explaining. A brain-computer interface bypasses damaged neural pathways, turns intentions straight into machine commands and builds a bridge between a person and a screen, giving expression and interaction an outlet again.
The company's website calls these participants pioneers and lists, one by one, the year of each implant and the underlying condition. According to media reports, some can already use thought to guide a robotic arm to pick up a cup, and others have steered a motorized wheelchair in tests; the device connects to more than a computer.
Twelve patients is not many, yet it is enough for researchers and regulators to gather real-world data. Every successful decode feeds back into the algorithms, and every surprise is a reminder that engineering still has gaps to close, which is why medicine became the first place this technology took root.
According to the Xinhua news app, the company splits its devices in development into several lines: some target movement disorders, helping people control external tools by thought; some aim at restoring vision; and others plan to send electrodes into deeper brain regions for neuromodulation. That clear division of labor is one footnote to medicine being the first setting to work.
A Thread Through the Dura Mater
What really drew the industry's attention was that the surgery itself got shorter. According to CCTV, the company announced that a single electrode thread could pass straight through the dura mater, the tough membrane wrapping the brain, with no need to cut it open, which means less trauma and a process easier to standardize.
The same report noted that insertion time for a single thread fell from 17 seconds to 1.5 seconds, and insertion depth passed 50 millimeters, reaching neurons deeper in the brain. A single operation often places hundreds or thousands of threads, so the second or so saved on each one adds up to a substantial gain in efficiency.
Manufacturing costs have come down too. According to CCTV and a brokerage research note, the cost of making the key components fell by about 95 percent, and the total cost of one device plus the surgery was pushed under US$50,000. The company says that once this threshold is crossed, scaling up has a realistic foundation.
One thing should be clear: that US$50,000 covers the device and the surgery together, and cannot simply be read as the sticker price of a machine. It is more like a payment threshold. Cross it and more patients can afford it; fail to cross it and brain-computer interfaces remain an option only a few can reach.
How the Costs Came Down
The savings did not come from one clever trick, but from handing the steps most dependent on a surgeon's touch over to machines. Removing part of the dura used to test a surgeon's experience; with an automated surgical robot in charge, variation between individuals shrinks, and replication becomes possible.
The supply chain is the other half. Only when core parts can be mass-produced like consumer electronics does marginal cost truly fall. That is why the factory logic keeps coming up: turning delicate hand work into a process that can be repeated.
But falling fast on cost is only one side of the ledger. Surgery, post-operative tuning, long-term follow-up and the payment system all have to be counted in. If any one of these is not caught, that handsome figure stays on paper and never becomes a routine procedure an ordinary family can afford.
According to Guangming Online, around 2026 the company spoke repeatedly of moving toward a leaner, more automated surgical workflow. Plans that push the timetable out and the case numbers already verifiable at the bedside sit on two quite different scales.
A Roadmap Still on Paper
Further off than the surgery is a plan stretching to 2028. According to CCTV and the Xinhua news app, the company aims to raise the electrode channel count to 25,000, extend implantation from the motor cortex to the speech and visual cortices, and support several devices working in parallel.
In the same blueprint, a vision-restoration system called Blindsight has received breakthrough device designation from the US Food and Drug Administration. The company has also set out the ultimate goal of a whole-brain interface, imagining a high-bandwidth link between the biological brain and external machines.
According to Xinhua and Guangming Online, integrating the human brain more closely with artificial intelligence is written down as a goal for a later stage, not something that has already happened. Telling what has landed from what is still in the plan is the single most important step in reading news like this.
For ordinary readers, what can be checked is the number of implants, the speed of insertion and the cost basis; what is still on paper is the channel count, the brain areas covered and full integration. Mistaking a vision for an accomplished fact is exactly the trap popular-science writing most easily falls into.
The Gap Between Vision and Line
In the public record, this company tends to announce a very broad goal first and then advance the engineering in stages. That kind of story draws attention and capital, but it also places the pressure of delivery squarely on every single process.
According to Yicai and other media reports, people in the field are generally cautious about the timetable for consumer implants, reckoning that years of validation still lie between regulatory approval for medical use and anything an ordinary person can use. Beyond the lively launch events, this road has to be paved one improvement at a time.
A production line answers to what can be repeated, verified and scaled; a vision answers to room for imagination, and the two were never measured with the same ruler. Telling both stories well is not hard; the hard part is keeping the latter from running so far ahead that it throws off the former's rhythm.
Getting Medicine Right First
Back to the picture at the start: a person moves a cursor by intention without lifting a hand. The most moving moment of this technology is always at the bedside; whether it can grow into a repeatable business depends on grinding the surgery, the cost and the follow-up down to something stable.
A vision has value because it points the way; a production line has more value because it alone acknowledges what has actually been done. For an ordinary reader watching from the sidelines, rather than memorizing a distant goal, it is better to watch whether the next group of patients and the next process really land.
Disclaimer: This article is based on publicly available information and is for reference only. It does not constitute investment advice.