Breakthrough Technologies 2026: MIT's Annual List

Overview

MIT Technology Review dropped its 25th annual list of breakthrough technologies on January 12, 2026—250 predictions over a quarter century. This year's ten picks span sodium-ion batteries poised to power the next generation of cheap EVs, generative AI that's rewriting how software gets built, and personalized CRISPR treatments custom-made for individual babies. The list includes embryo screening for intelligence that's reigniting eugenics debates and hyperscale data centers devouring city-sized power loads to train AI models.

The stakes are high. MIT Tech Review called natural-language processing in 2001, CRISPR in 2015, and mRNA vaccines in 2017—technologies that became multi-trillion-dollar industries and reshaped society. But they've also bet wrong: social TV flopped, Apple Vision Pro tanked, and robotaxis crashed. The 2026 list signals where industry leaders are placing their chips, from CATL's bet on sodium batteries hitting production by year's end to Fyodor Urnov launching a startup to mass-produce bespoke gene therapies.

Play on this story Voices Debate Predict

Key Indicators

Years of predictions

250 technologies identified as breakthroughs since 2001

90 TWh

AI data center power by 2026

Tenfold increase from 2022 levels

175 Wh/kg

Sodium-ion battery density

CATL's Naxtra platform achieving commercial viability

6 months

Baby KJ treatment development

From diagnosis to personalized CRISPR therapy

Interactive

Exploring all sides of a story is often best achieved with Play.

Ever wondered what historical figures would say about today's headlines?

People Involved

Amy Nordrum

Led the 2026 breakthrough technologies selection

Robin Zeng Yuqun

Launching sodium-ion battery production at scale in 2026

Fyodor Urnov

Scaling personalized CRISPR medicines following Baby KJ success

Dario Amodei

Leading mechanistic interpretability research to understand AI models

Organizations Involved

MIT Technology Review

Technology Publication

Published 25th annual breakthrough technologies list

The definitive voice on emerging technologies with a 125-year track record.

Contemporary Amperex Technology Co. Limited (CATL)

Battery Manufacturer

Launching mass production of sodium-ion batteries in 2026

World's largest EV battery manufacturer pioneering sodium-ion technology.

Aurora Therapeutics

Biotechnology Startup

Developing personalized CRISPR therapies using FDA's plausible mechanism pathway

Startup scaling personalized CRISPR medicines following Baby KJ breakthrough.

Anthropic

AI model developer

Leading mechanistic interpretability research to understand AI models

AI safety company racing to understand models before they become superintelligent.

American Society for Reproductive Medicine (ASRM)

Professional Medical Organization

Concluded polygenic embryo screening not ready for clinical use

Medical society declaring embryo intelligence screening scientifically premature and ethically fraught.

Timeline

MIT Technology Review Publishes 2026 Breakthrough Technologies List
Publication

25th annual list identifies sodium-ion batteries, generative coding, base-edited therapies, mechanistic interpretability, hyperscale data centers, AI companions, gene resurrection, commercial space stations, embryo scoring, and new nuclear reactors.
January 12th, 2026
Aurora Therapeutics Launches
Company Launch

Doudna and Urnov launched Aurora with $16M to scale personalized CRISPR medicines.
January 9th, 2026
CATL Confirms 2026 Sodium Battery Production
Announcement

At supplier conference, confirmed mass production of Naxtra sodium-ion batteries by late 2026.
December 28th, 2025
ASRM Declares Embryo Screening Not Ready
Professional Guidance

Medical society concluded polygenic embryo screening should not be offered clinically.
December 1st, 2025
California Passes AI Chatbot Safety Law
Regulation

SB 243 signed requiring protections for minors using AI chatbots, effective January 2026.
October 13th, 2025
Baby KJ Discharged Home After Successful Treatment
Medical Milestone

After 10 months hospitalized, KJ went home following three CRISPR infusions.
June 1st, 2025
Anthropic Sets 2027 AI Interpretability Goal
Announcement

Dario Amodei announced goal to reliably detect most AI model problems by 2027.
April 24th, 2025
Baby KJ Receives First Personalized CRISPR Therapy
Medical Treatment

Seven-month-old KJ received bespoke base editing therapy for CPS1 deficiency.
February 1st, 2025
CRISPR Treatment for Sickle-Cell Featured
Publication

First gene-editing treatment earned regulatory approval, validating 2015 prediction.
January 8th, 2024
CATL Unveils First Sodium-Ion Battery
Product Launch

Robin Zeng unveiled first-generation sodium-ion battery with 160 Wh/kg density.
July 1st, 2021
mRNA Vaccine Platforms Featured on List
Publication

MIT Tech Review highlighted mRNA vaccine platforms years before COVID-19 pandemic.
January 1st, 2017
CRISPR Identified as Breakthrough Technology
Publication

MIT Tech Review called CRISPR gene editing the biggest biotech breakthrough of the century.
January 1st, 2015
First Breakthrough Technologies List Published
Publication

MIT Technology Review published its inaugural list, identifying natural-language processing as a breakthrough technology.
January 1st, 2001

Scenarios

Predict which scenario wins. Contrarian picks score more — points lock in when the scenario resolves.

Predictions leaderboard

Sodium-Ion Disrupts Lithium Dominance by 2030

CATL's Naxtra platform enters mass production in late 2026 and captures significant market share for grid storage and affordable EVs by 2028. The technology's cheaper materials, safer operation, and wider temperature range make it ideal for stationary storage and budget vehicles. By 2030, sodium-ion batteries claim half the lithium iron phosphate market as Zeng predicted, reducing dependence on lithium mining and enabling sub-$20,000 EVs in developing markets. BYD's 30 GWh plant in Xuzhou becomes operational, and Chinese manufacturers export sodium-ion technology globally, fundamentally reshaping the energy storage landscape.

Discussed by: CATL chairman Robin Zeng, MIT Technology Review editors, battery industry analysts

Consensus —

Personalized CRISPR Therapies Become Standard Care

Aurora Therapeutics successfully navigates FDA's plausible mechanism pathway in 2027, gaining approval for multiple phenylketonuria therapies targeting different mutations. Baby KJ's continued health provides long-term safety data that accelerates regulatory acceptance. By 2028, other biotechs adopt the model, and insurance begins covering personalized gene therapies for ultra-rare diseases. The six-month development timeline that worked for Baby KJ becomes standardized, and bespoke CRISPR treatments expand from metabolic disorders to cancer and autoimmune conditions. What was impossible in 2024 becomes routine by 2030, though only for patients in wealthy countries with advanced medical infrastructure.

Discussed by: Aurora Therapeutics, FDA Commissioner Marty Makary, biotech analysts, Nature

Consensus —

AI Interpretability Fails, Models Become Undeployable

Anthropic and other labs make progress on mechanistic interpretability but can't keep pace with capability improvements. By late 2027, AI systems reach the country of geniuses threshold Amodei warned about, but interpretability techniques still can't reliably detect deception, goal misalignment, or emergent behaviors. A high-profile AI failure—possibly in autonomous weapons, financial systems, or critical infrastructure—causes governments to impose deployment freezes on frontier models until interpretability catches up. The AI boom stalls as companies can't ship products they can't explain or control, validating Amodei's warning that the race between understanding and capability would determine whether advanced AI gets deployed safely.

Discussed by: Dario Amodei, AI safety researchers, tech policy analysts

Consensus —

Breakthrough List Predictions Prove Mixed, Like Past 25 Years

The 2026 list follows the historical pattern: some technologies become massive industries, others remain niche, and a few flop entirely. Sodium-ion batteries and generative coding succeed commercially but don't transform society as dramatically as predicted. Personalized CRISPR therapies help thousands but remain too expensive for widespread adoption. Embryo intelligence screening triggers regulatory crackdowns before achieving scale. Commercial space stations launch but serve tiny markets. The list's real value continues to be identifying emerging technologies worth watching, not predicting exact timelines or market sizes. Like natural-language processing in 2001 or CRISPR in 2015, the 2026 picks prove directionally correct but take longer than expected to mature.

Discussed by: MIT Technology Review editors, technology historians, Gartner analysts

Consensus —

Historical Context

Gartner Hype Cycle Technology Forecasting (1995-Present)

1995-2026

What Happened

Gartner introduced the Hype Cycle framework in 1995 to predict technology adoption through five phases: innovation trigger, peak of inflated expectations, trough of disillusionment, slope of enlightenment, and plateau of productivity. The framework became widely referenced in enterprise technology but faced substantial criticism for lack of scientific rigor. Research analyzing Gartner Hype Cycles since 2000 found that few technologies actually travel through an identifiable hype cycle, and most important technologies adopted since 2000 weren't identified early in their adoption cycles. Industry analyst Duncan Stewart noted the cycle is a lousy predictor—some technologies like cloud computing and 4G wireless surpassed the hype, others like 3D printing followed the curve almost perfectly, but being at hype peak doesn't guarantee success or failure.

Outcome

Short Term

Hype Cycle became standard enterprise framework despite accuracy concerns.

Long Term

Framework proved better at describing technology psychology than predicting outcomes; only one-fifth of hyped technologies followed expected adoption pattern.

Why It's Relevant Today

Shows the difficulty of technology forecasting even with systematic frameworks—MIT Tech Review's 25-year track record faces similar challenges distinguishing hype from genuine breakthroughs.

World Economic Forum Top 10 Emerging Technologies (2012-Present)

2012-2026

What Happened

The World Economic Forum partnered with Frontiers to produce annual lists of emerging technologies expected to deliver impact within three to five years. The report pools expertise from hundreds of leading scientists worldwide through a multi-phase methodology combining peer-reviewed evidence, citation analysis, and expert consensus. The WEF's track record includes identifying CRISPR in 2015, mRNA vaccine platforms in 2017, and AI in drug discovery in 2020—all of which proved transformative. The 2025 list featured generative AI watermarking, collaborative sensing, structural battery composites, and engineered living therapeutics. Unlike Gartner's descriptive framework, WEF focuses on scientific readiness and societal impact across health, sustainability, safety, and industrial applications.

Outcome

Short Term

WEF correctly identified CRISPR, mRNA vaccines, and AI drug discovery before mainstream adoption.

Long Term

Three-to-five-year timeframe proved accurate for many predictions; scientific community validation added credibility compared to analyst-driven forecasts.

Why It's Relevant Today

Provides validation model for MIT Tech Review's approach—both rely on expert consensus and scientific readiness rather than market hype, and both have strong track records predicting CRISPR and mRNA technologies.

MIT Tech Review's Past Predictions: Natural Language Processing (2001)

2001-2026

What Happened

In the inaugural 2001 Breakthrough Technologies list, MIT Tech Review featured natural-language processing, highlighting Microsoft Research's Karen Jensen and the MindNet system for automatically extracting concept networks from dictionaries. The article envisioned extended conversations with computers and video mining using combined speech recognition, image understanding, and NLP techniques. At the time, there was already commercial speech recognition for dictation and software that could understand plain-English queries. The vision was to create interfaces allowing natural human-computer interaction, with HAL 9000 from 2001: A Space Odyssey serving as cultural touchstone for what was possible.

Outcome

Short Term

NLP remained primarily in research labs through the 2000s with incremental commercial applications in search and voice assistants.

Long Term

By 2026, NLP became foundational to trillion-dollar industries: search engines, virtual assistants (Siri, Alexa), translation services, and large language models like GPT-4 and Claude that fulfill the conversational AI vision from 2001.

Why It's Relevant Today

Demonstrates MIT Tech Review's 25-year foresight—NLP seemed futuristic in 2001 but became ubiquitous infrastructure by 2026, validating the list's ability to identify technologies before they reach mainstream awareness.