News
Scientists use AI neural network to translate speech from brain activity
Three recently published studies focused on using artificial intelligence (AI) neural networks to generate audio output from brain signals have shown promising results, namely by producing identifiable sounds up to 80% of the time. Participants in the studies first had their brain signals measured while they were either reading aloud or listening to specific words. All the data was then given to a neural network to “learn” how to interpret brain signals after which the final sounds were reconstructed for listeners to identify. These results represent hopeful prospects for the field of brain-computer interfaces (BCIs), where thought-based communication is quickly moving from the realm of science fiction to reality.
The idea of connecting human brains to computers is far from new. In fact, several relevant milestones have been made in recent years including enabling paralyzed individuals to operate tablet computers with their brain waves. Elon Musk has also famously brought attention to the field with Neuralink, his BCI company that essentially hopes to merge human consciousness with the power of the Internet. As brain-computer interface technology expands and develops new ways to foster communication between brains and machines, studies like these, originally highlighted by Science Magazine, will continue demonstrating the steady march of progress.
In the first study conducted by researchers from Columbia University and Hofstra Northwell School of Medicine, both in New York, five epileptic participants had the brain signals from their auditory cortexes recorded as they listened to stories and numbers being read to them. The signal data was provided to a neural network for analysis which then reconstructed audio files that were accurately identified by participating listeners 75% of the time.
In the second study conducted by a team from the University of Bremen (Germany), Maastricht University (Netherlands), Northwestern University (Illinois), and Virginia Commonwealth University (Virginia), brain signal data was gathered from six patients’ speech planning and motor areas while undergoing tumor surgeries. Each patient read specific words aloud to target the data collected. After the brain data and audio data were given to their neural network for training, the program was given brain signals not included in the training set to recreate audio, the result producing words that were recognizable 40% of the time.
Finally, in a third study by a team at the University of California, San Francisco, three participants with epilepsy read text aloud while brain activity was captured from the speech and motor areas of their brains. The audio generated from their neural network’s analysis of the signal readings was presented to a group of 166 people who were asked to identify the sentences from a multiple choice test – some sentences were identified with 80% accuracy.
While the research presented in these studies shows serious progress towards connecting human brains to computers, there are still a few significant hurdles. For one, the way neuron signal patterns in the brain translate into sounds varies from person to person, so neural networks must be trained on each individual person. The best results require the best data possible, i.e., the most precise neuron signals possible, meaning this is something that can only be obtained by placing electrodes in the brain itself. The opportunities to collect data at this invasive level for research are limited, relying on voluntary participation and approval of experiments.
All three of the studies highlighted demonstrated an ability to reconstruct speech based on neural data in some significant capacity; however, also in all cases, the study participants were able to create audible sounds to use with the computer training set. In the case of patients unable to speak, the level of difficultly in interpreting the brain’s speech signals from other signals will be the biggest challenge. Also, the differences between brain signals during actual speech vs. thinking about speech will complicate matters further.
News
Tesla gives HW3 owners another massive update
It was an “at last” moment for HW 3 owners, who have waited for an update on the capabilities of their vehicles for some time. After CEO Elon Musk finally admitted last week that the HW3 vehicles would not be capable of unsupervised FSD, it appears Tesla is bringing a new, more transparent tone to those owners.
Tesla is giving Hardware 3 vehicle owners another massive update, the second major communication the company has given to those drivers after what seemed like years of being left out to dry.
The company, which plans to launch a Full Self-Driving version 14 iteration that is compatible with these cars, which have older chips, is now planning to expand the rollout of the v14 Lite offering to other markets, it said on X.
Tesla said:
“Following future rollout of FSD V14 Lite for HW3 vehicles in the US, we plan on expanding V14 Lite to additional international markets. This update ensures that HW3 vehicle owners will continue to benefit from ongoing software updates. Since international rollout is subject to several factors (completion of technical verification, regional adaptation & relevant regulatory approvals), we can’t provide definitive dates at the moment, but will provide updates on a rolling basis.”
This announcement comes at a critical time for HW3 owners, many of whom purchased Full Self-Driving (FSD) capability years ago with promises of ongoing support and future-proofing.
Following future rollout of FSD V14 Lite for HW3 vehicles in the US, we plan on expanding V14 Lite to additional international markets.
This update ensures that HW3 vehicle owners will continue to benefit from ongoing software updates.
Since international rollout is subject to…
— Tesla (@Tesla) April 29, 2026
HW3, introduced in 2019, powers vehicles from roughly 2019 to early 2023 models. While newer AI4 hardware has advanced rapidly, HW3 owners have felt increasingly left behind, with their last major update stuck around version 12.6 since early 2025.
It was an “at last” moment for HW 3 owners, who have waited for an update on the capabilities of their vehicles for some time. After CEO Elon Musk finally admitted last week that the HW3 vehicles would not be capable of unsupervised FSD, it appears Tesla is bringing a new, more transparent tone to those owners.
V14 Lite represents a significant optimization effort. Tesla has confirmed it will bring many core features of the full V14 release, currently running on more powerful hardware, to the more constrained HW3 platform.
Expected capabilities include improved handling of complex urban scenarios, better reverse driving, enhanced parking features, and smoother overall autonomy, albeit in a “lite” form tailored to HW3’s compute limits. Tesla’s head of Autopilot, Ashok Elluswamy, noted during the Q1 2026 earnings call that the update is targeted for late June in the U.S.
Tesla is releasing a modified version of FSD v14 for Hardware 3 owners: here’s when
The international expansion is particularly meaningful for owners in Europe, Asia, Australia, and other regions where FSD rollout has lagged due to regulatory hurdles.
Tesla emphasized that timing remains fluid, dependent on “technical verification, regional adaptation & relevant regulatory approvals.” No firm dates were provided, but the company pledged rolling updates as milestones are achieved.
This move addresses growing concerns that Tesla might abandon legacy hardware. With the recent admission that its capabilities are limited and not capable of Tesla’s grand autonomy ambitions, owners are finally in the light of truth, with more honesty being put forth as the company navigates this chapter.
For Tesla, keeping HW3 relevant strengthens customer loyalty and protects the value of older vehicles. It also buys time as the company pushes toward broader regulatory approvals and unsupervised autonomy on newer platforms.
While V14 Lite isn’t the full unsupervised experience once promised, it delivers tangible improvements and signals that HW3 owners are not being forgotten.
As Tesla continues its rapid AI and autonomy evolution, this update underscores a key principle: software can breathe new life into existing hardware. For tens of thousands of HW3 drivers worldwide, V14 Lite could mark the beginning of a renewed era of confidence in their vehicles.
Elon Musk
SpaceX Board has set a Mars bonus for Elon Musk
SpaceX has given Elon Musk the goal to put one million people on Mars.
SpaceX’s board approved a compensation plan for Elon Musk that ties his pay directly to colonizing Mars and building data centers in outer space. The details surfaced this week after Reuters reviewed SpaceX’s confidential registration statement filed with the Securities and Exchange Commission, making it one of the first concrete looks inside the company’s financials ahead of a public offering.
The pay package will reportedly award Musk 200 million super-voting restricted shares if the company hits a market valuation milestone, with the most ambitious targets going further. To unlock the full award, SpaceX would need to reach a $7.5 trillion valuation and help establish a permanent human settlement on Mars with at least one million residents. Additional incentives are tied to developing space-based computing infrastructure capable of delivering at least 100 terawatts of processing power.
SpaceX wins its first MARS contract but it comes with a catch
Long before SpaceX filed anything with the SEC, Elon Musk had already spent years framing Mars colonization as an insurance policy against human extinction. The philosophy traces back to at least 2001, when Musk first began researching Mars missions independently, before SpaceX even existed. By 2002 he had founded the company with Mars as the stated long-term goal.
In a 2017 presentation at the International Astronautical Congress, Musk outlined the specific vision that still underpins SpaceX’s architecture today. He described a self-sustaining city on Mars requiring roughly one million people to become viable, the same number now written into his compensation package.
SpaceX’s Starship, still in active development, was designed from the ground up to support the eventual colonization of Mars. Musk has stated publicly that getting the cost per ton to Mars below $100,000 is necessary to make mass migration economically feasible. Everything from Starship’s payload capacity to its full reusability targets flows from that single constraint. One can say that Musk’s latest compensation package has put a formal valuation on Mars for the first time.
SpaceX is targeting an IPO around June 28, Musk’s birthday, at a valuation of approximately $1.75 trillion. Between the Mars rover contract, the Golden Dome software group, Space Force satellite launches, and now a pay structure built around interplanetary colonization, SpaceX has become the single most consequential contractor in American space and defense. The IPO will put a public price tag on all of it for the first time.
Earlier this week, we wrote a story on how Tesla is launching a new Supercharging Queue system to mitigate problems between drivers when there is a wait to charge.
Rather than potentially having people end up in a physical conflict, Tesla’s approach is to determine who is next to charge based on geographic data.
Tesla launches solution to end Supercharger fights once and for all
But some companies, notably Tesla’s biggest rival in China, BYD, are taking a different approach, focusing on charging speeds rather than how they will manage delays.
BYD’s approach, especially with its tests of ultra-fast “Flash Charging” technology, is to eliminate the length of a charging session. At the heart of this strategy is BYD’s second-generation Blade Battery paired with 1,500-kW Flash Chargers.
Real-world FLASH Charging in action.
⚡ 10% → 70% in 5 minutes
⚡ 10% → 97% in 9 minutesIntroducing BYD’s 2nd Generation Blade Battery + FLASH Charging Technology.
20,000 stations will bring faster, safer, and smarter EV charging across China by the end of 2026. pic.twitter.com/uzQC8q1xGf
— BYD (@BYDCompany) March 9, 2026
Unveiled earlier this year, the system charges compatible vehicles from 10 percent to 70 percent state of charge in just five minutes and from 10 percent to 97 percent in nine minutes.
Real-world demonstrations on models like the Yangwang U7 and Denza Z9 GT have shown the tech delivering roughly 250 miles (400 kilometers) of range in just five minutes. This would essentially match or beat the time it takes to fill a gas tank.
Sometimes, gas pumps get congested, and there are lines. You rarely see conflicts at pumps because filling up a tank rarely takes more than five minutes.
Tesla’s fastest Supercharger build currently is the v4, which can deliver up to 325 kW for Cybertruck and 250 kW for other models, but there are “true” sites that are capable of up to 500 kW. This enables speeds of up to 1,000 miles per hour, or 1,400 miles for 350 kW-capable vehicles.
The breakthrough stems from BYD’s vertically integrated ecosystem: a new 1,000-volt architecture, 10C charging rates, and proprietary silicon-carbide chips that minimize internal resistance while protecting battery health.
The company plans to install 20,000 Flash Charging stations across China by the end of 2026, with thousands already operational and global expansion eyed for Europe and beyond later this year.
Early rollout targets popular models, including upgrades to high-volume sellers like the Seal and Sealion series, bringing five-minute charging to mainstream prices around 100,000 yuan (about $14,000).
This approach contrasts sharply with Tesla’s software solution. Tesla’s Virtual Queue uses geofencing and the app to assign turns at crowded sites, addressing driver disputes and idle time. It’s a clever fix for today’s network realities.
Yet, BYD’s philosophy is simpler: make charging so fast that waits barely exist. A five-minute stop becomes as convenient as a gas-station visit, reducing station dwell time, easing grid strain, and lowering range anxiety for long trips.
For consumers, the difference is potentially tangible. They’ll spend more time driving and less time parked. It is just another way Tesla and BYD are pushing one another to improve the overall experience of EV ownership.
Tesla gives HW3 owners another massive update
SpaceX Board has set a Mars bonus for Elon Musk
