0:00
/

The great BCI bubble of the 2020 why permanent implants are already dead and why fully Resorbable Systems are the only path forward

A Deep Dive into Hype, Biology, Regulation, Ethics, Economics, and the Coming AI-Neural Risks

The brain-computer interface (BCI) sector stands at a fever pitch in early 2026. Venture capital has poured in more than $4 billion over the past five years. Paper valuations for leading players exceed $50 billion combined. Elon Musk’s Neuralink continues raising at sky-high multiples, Synchron touts FDA Breakthrough Device Designation, Paradromics and Precision Neuroscience push aggressive timelines, and dozens of startups chase the dream of merging human minds with silicon.

Yet the clinical reality is sobering. No high-channel-count BCI has achieved broad commercial approval. Implants remain confined to tiny investigational cohorts—mostly patients with severe paralysis from ALS or spinal injury. Performance degrades over time. Complications persist. And the grand promises of consumer telepathy, photorealistic vision restoration, or seamless AI symbiosis feel increasingly detached from physics, biology, and the collapsing realities of modern healthcare.

Max Hodak, Neuralink’s co-founder and former president, left the company in 2021. Since then, he has offered some of the field’s most candid assessments. No longer bound by corporate messaging, his warnings read like a postmortem on the permanent-implant paradigm.

Max Hodak’s Post-Neuralink Reality Check (2023–2025)

Hodak’s public statements cut through the hype:

•  On timelines: “Consumer-scale deployment in the 2020s is effectively impossible unless the device dissolves or is trivially explantable.”

•  On chronic stability: “Every material we tried gliosed over in primates within 12–18 months. Chronic high-channel recording without constant surgical revision is not solvable.”

•  On bandwidth: “10,000 channels sounds impressive until you do the math… you’re lucky to get 200 reliable neurons. That is ~150 bits per second on a good day. Anyone telling you that’s going to give you photorealistic vision or natural speech is selling science fiction.”

•  On patient acceptance: “No healthy human being is going to volunteer to have something sewn into their cortex for the rest of their life when the incremental benefit over a noninvasive or temporary system is marginal.”

•  On the bubble: “This is going to be the biggest biotech bubble since CAR-T in 2017 or gene therapy in 1999… The smart money is already rotating into resorbables and non-invasive.”

These are not the words of an outsider. They come from the man who helped design Neuralink’s early architecture, surgical robot, and vision. When he says the core premise of lifelong, high-bandwidth permanent implants is biologically unsound, investors and founders should listen.

Neuralink N1 implant with its ultra-thin flexible threads, shown next to a quarter for scale. Impressive engineering—yet still vulnerable to the biological realities Hodak describes.

The Biology That Refuses to Cooperate: Gliosis and Signal Decay

Every permanent implant—rigid Utah arrays, flexible threads, stent-based systems—triggers the brain’s foreign-body response. Microglia activate, astrocytes form a glial scar, neurons retract or die. Signal quality plummets 50–90% within months to years.

Histology of glial scarring around implanted electrodes. The bright green and red show reactive astrocytes and microglia encapsulating the probe—classic foreign-body response.

This is not a solvable “coating problem.” It is fundamental immunology. The brain treats persistent hardware like an infection that must be walled off. Hodak saw this in primates. Every prior generation (BrainGate, Blackrock, etc.) has published the same pattern.

Progressive gliosis diagram over weeks to months, showing encapsulation that kills chronic performance.

Head-to-Head Reality Check: Current Systems vs. Outcomes (Early 2026)

Synchron Stentrode — endovascular approach reduces craniotomy but sacrifices bandwidth and introduces vascular risks.

Patient Experiences: Impressive Yet Limited and Fragile

Trial participants have achieved cursor control, slow typing, and basic gaming. Noland Arbaugh (Neuralink) and others demonstrate real progress for locked-in patients. Yet signals degrade, revisions loom, and the psychological burden of permanent hardware grows.

These are heroic n=1 stories, not scalable therapies. Performance often drops within a year.

The Four (Plus Two) Walls Crushing Permanent Implants

1.  Foreign-Body Response — Inevitable gliosis.

2.  Lifelong Recommitment — Patients and payers locked into open-ended risk.

3.  Regulatory Timeline — Class III permanent devices demand 10–20+ years of data.

FDA approval process for high-risk implants is long and rigorous—PMA pathway with extensive safety data.

4.  Bandwidth Ceiling — Real useful throughput remains tiny.

5.  Ethics of Neural Information — Permanent systems create lifelong data streams. Who owns decoded thoughts? Consent for what? Risks of “mind reading,” personality alteration, or coercion.

6.  Funding & Scale in a Collapsing Healthcare System — Procedures could cost $100k–$350k+ plus lifelong follow-up. U.S. healthcare costs rise 7–8% annually amid strained Medicare/Medicaid and payer pushback. Value-based care demands clear, time-limited outcomes. A system nearing insolvency cannot subsidize widespread lifetime brain hardware.

Resorbables collapse these problems. The device performs its job (weeks to months), then hydrolyzes harmlessly. No chronic scar, no explant surgery, shorter regulatory path (akin to bioabsorbable stents), defined consent windows, and far lower total cost of care.

FDA approval process for high-risk implants is long and rigorous—PMA pathway with extensive safety data.

4.  Bandwidth Ceiling — Real useful throughput remains tiny.

5.  Ethics of Neural Information — Permanent systems create lifelong data streams. Who owns decoded thoughts? Consent for what? Risks of “mind reading,” personality alteration, or coercion.

6.  Funding & Scale in a Collapsing Healthcare System — Procedures could cost $100k–$350k+ plus lifelong follow-up. U.S. healthcare costs rise 7–8% annually amid strained Medicare/Medicaid and payer pushback. Value-based care demands clear, time-limited outcomes. A system nearing insolvency cannot subsidize widespread lifetime brain hardware.

Resorbables collapse these problems. The device performs its job (weeks to months), then hydrolyzes harmlessly. No chronic scar, no explant surgery, shorter regulatory path (akin to bioabsorbable stents), defined consent windows, and far lower total cost of care.

Resorbable examples: Biodegradable wireless sensors and flexible interfaces that dissolve after therapy.

Self-deploying ECoG arrays that unfold then biodegrade.

The AI-Integration Nightmare: Misinformation, Thought Control, and Narrative Dominance

The permanent-implant dream becomes far darker when fused with advanced AI. Real-time, high-fidelity neural readout + generative AI creates bidirectional loops capable of:

•  Decoding and shaping thoughts in real time. AI could detect “undesirable” patterns (political dissent, emotional states) and intervene with stimulation or suggestions.

•  Narrative control at scale. State or corporate actors with access could push personalized misinformation directly into cognition, blurring the line between internal thought and external influence.

•  Misinformation amplification. Decoded intentions fed into LLMs could generate hyper-personalized propaganda or deepfakes that feel like one’s own ideas.

•  Thought control risks. Closed-loop systems for “mental health” could suppress nonconforming beliefs under the guise of therapy. Cognitive liberty—the right to one’s own unfiltered mind—becomes negotiable.

Conceptual illustration of AI-BCI fusion: brain linked to vast neural networks, raising profound questions of autonomy.

Temporary resorbables mitigate this. Limited duration means limited exposure. Data streams have hard expiration dates. Consent is time-bound. The ethical and societal risks shrink dramatically when the hardware does not outlive its therapeutic purpose.

Neurorights frameworks (proposed in Chile, Spain, and U.S. states) remain underdeveloped. Permanent systems exacerbate gaps; resorbables buy time for society to catch up.

Real Markets: Healthcare Outcomes at Scale

These applications serve millions today. They do not require sci-fi bandwidth or lifetime commitment. They require reliable, safe, reimbursable therapy windows.

Why the Bubble Will Pop (2026–2028)

Valuations assume consumer-scale, high-bandwidth, permanent implants. Biology, regulators, payers, ethicists, and patients say no. Hodak’s lived experience inside the best-funded attempt confirms it.

When the first major adverse event hits headlines—or when reimbursement denials cascade—the capital flight will be swift. Smart money is already shifting toward resorbables, non-invasive hybrids, and pragmatic medical applications.

The most ridiculous statement heard From the Chief of staff of an endovascular BCI company I heard was he what to do a clinical study with no endpoint to use as “retrospective data for approval

Conclusion: The Final Paradigm Is Healthcare — Not Bandwidth, Valuation, or Sci-Fi Control

The BCI bubble is not bursting solely because bandwidth dreams failed or valuations detached from reality. It is collapsing under the weight of biological incompatibility, regulatory rigor, ethical minefields (especially with AI integration enabling thought and narrative control), and economic unsustainability in a strained healthcare system.

Permanent implants wage endless war on the brain. Resorbables work with it—delivering intense, high-fidelity access exactly when needed, then vanishing without trace.

Max Hodak tried to build the permanent future and walked away convinced it cannot scale. The survivors will be those who accepted biology’s terms early: temporary, therapeutic, ethical, affordable interfaces that improve real patient outcomes.

The era of lifetime hardware in the brain is ending. The era of responsible, healthcare-grounded neural medicine is beginning. When the dust settles, the winners won’t be those promising to upload humanity to the cloud. They will be the ones who quietly, safely, and effectively helped patients recover, manage disease, and reclaim function—without chaining them to forever foreign bodies or opening the door to dystopian mind control.

Resorbables aren’t the compromise.

They are the only architecture aligned with human biology, regulatory reality, ethical guardrails, payer constraints, and—most importantly—patients in the real world.

The future of BCIs isn’t about merging with AI at any cost.

It’s about better healthcare, delivered responsibly. That future is already unfolding in quiet labs working on devices designed to disappear when their job is done.

Discussion about this video

User's avatar

Ready for more?