News
Google’s DeepMind unit develops AI that predicts 3D layouts from partial images
Google’s DeepMind unit, the same division that created AlphaGo, an AI that outplayed the best Go player in the world, has created a neural network capable of rendering an accurate 3D environment from just a few still images, filling in the gaps with an AI form of perceptual intuition.
According to Google’s official DeepMind blog, the goal of its recent AI project is to make neural networks easier and simpler to train. Today’s most advanced AI-powered visual recognition systems are trained through the use of large datasets comprised of images that are human-annotated. This makes training a very tedious, lengthy, and expensive process, as every aspect of every object in each scene in the dataset has to be labeled by a person.
The DeepMind team’s new AI, dubbed the Generative Query Network (GQN) is designed to remove this dependency on human-annotated data, as the GQN is designed to infer a space’s three-dimensional layout and features despite being provided with only partial images of a space.
Similar to babies and animals, DeepMind’s GQN learns by making observations of the world around it. By doing so, DeepMind’s new AI learns about plausible scenes and their geometrical properties even without human labeling. The GQN is comprised of two parts — a representation network that produces a vector describing a scene and a generation network that “imagines” the scene from a previously unobserved viewpoint. So far, the results of DeepMind’s training for the AI have been encouraging, with the GQN being able to create representations of objects and rooms based on just a single image.
As noted by the DeepMind team, however, the training methods that have been used for the development of the GQN are still limited compared to traditional computer vision techniques. The AI creators, however, remain optimistic that as new sources of data become available and as improvements in hardware get introduced, the applications for the GQN framework could move over to higher-resolution images of real-world scenes. Ultimately, the DeepMind team believes that the GQN could be a useful system in technologies such as augmented reality and self-driving vehicles by giving them a form of perceptual intuition – extremely desirable for companies focused on autonomy, like Tesla.
Google DeepMind’s GQN AI in action. [Credit: Google DeepMind]
In a talk at Train AI 2018 last May, Tesla’s head of AI Andrej Karpathy discussed the challenges involved in training the company’s Autopilot system. Tesla trains Autopilot by feeding the system with massive data sets from the company’s fleet of vehicles. This data is collected through means such as Shadow Mode, which allows the company to gather statistical data to show false positives and false negatives of Autopilot software.
During his talk, Karpathy discussed how features such as blinker detection become challenging for Tesla’s neural network to learn, considering that vehicles on the road have their turn signals off most of the time and blinkers have a high variability from one car brand to another. Karpathy also discussed how Tesla has transitioned a huge portion of its AI team to labeling roles, doing the human annotation that Google DeepMind explicitly wants to avoid with the GQN.
Musk also mentioned that its upcoming all-electric supercar — the next-generation Tesla Roadster — would feature an “Augmented Mode” that would enhance drivers’ capability to operate the high-performance vehicle. With Tesla’s flagship supercar seemingly set on embracing AR technology, the emergence of new techniques for training AI such as Google DeepMind’s GQN would be a perfect fit for the next generation of vehicles about to enter the automotive market.
-
Tesla flexed the new features of the Model Y L with a dedicated media event at the company’s Diner on Santa Monica Boulevard in Los Angeles.
The Model Y L is the extended-wheelbase version of the all-electric crossover, which has been voted the best-selling car in the world on three occasions. The vehicle is already rolling off production lines at Gigafactory Texas, and first deliveries are slated to take place later this year.
Tesla brings Model Y L ‘Launch Series’ to the U.S. at $61,990
Teslarati was invited to the event, but due to some scheduling conflicts, we could not make it to Los Angeles. Instead, we will have our hands on a media unit sometime in August, so we’ll be able to spend some more extended time with the Model Y L.
However, plenty of those who made it to LA shared some cool features that set the Model Y L apart from the Model Y.
Multi-Row Climate Control
Tesla fitted the Model Y L with full climate control on all three rows on the front screen. It can be adjusted by selecting which row you’d like to modify on the right-hand side of the touch screen:
The Model Y L has a little row of icons on the right side of the climate screen so the front passenger can more quickly jump to whatever row they want to control. pic.twitter.com/uJS44mgCvk
— Sawyer Merritt (@SawyerMerritt) July 30, 2026
-
Better Rear Window Visibility
One of the strangest things about the Model Y, especially the Juniper iteration, is the rear window has extremely limited visibility when looking into the rearview mirror.
Tesla has improved upon this with the Model Y L:
This is a drastic difference. https://t.co/wTdhitD5aA pic.twitter.com/z6hFs0Jkss
— TESLARATI (@Teslarati) July 30, 2026
PowerShare will be included
Model Y L will come with PowerShare in North America, with an 11.5kW output to your home. Tesla said it would require Powerwall 3 for operation.
Wireless Charging Pad
There has been some speculation that Tesla would upgrade the wireless charging pads in the United States, but this is not the case.
-
Tesla owner fixes common feature complaint with crafty DIY retrofit
Tesla confirmed that it does not come with the cooled pads as the Y L in China does. This is because North America has not adopted Qi charging yet.
Thermal Management Improvements
These improvements in the Model Y L were seen with thermal management:
- Up to 15% faster cabin cooling
- +23% thermal efficiency gained in hot weather, 7 miles of real-world range gained
- 8x more solar energy reflection off of glass roof
- 30% reduction in solar energy entering the cabin
Some of the thermal management improvements in the new Model Y L Premium:
• Up to 15% faster cabin cooling
• Up to 15 min faster cabin cooling in sunny conditions
• +23% thermal efficiency gained in hot weather +7 miles of real-world range gained
• +10 miles recovered after… pic.twitter.com/jZirFWFdtO— Sawyer Merritt (@SawyerMerritt) July 31, 2026
Demand
Tesla said the Model Y L is almost sold out in the U.S. It comes with
-
- 1 year of FSD Supervised
- 1 year of Supercharging
- 1 year of Premium Connectivity
- Free exterior paint color, interior, and wheel option at no additional cost
Tesla CEO Elon Musk took to his social media platform X on Thursday night to deny a ridiculous rumor regarding the sale of the company’s Chinese vehicle production plant, Gigafactory Shanghai.
On Thursday, the Wall Street Journal, citing sources familiar with the matter, claimed in a scathing new report that Tesla was exploring a potential sale of the entire China business in an effort to help bolster a potential merger between SpaceX and Tesla.
Musk immediately denied the rumor not once but twice, initially calling it “fake news,” and then calling it “absurdly fake news” in a separate post just a few moments later:
This is fake news
— Elon Musk (@elonmusk) July 31, 2026
The original poster of the Wall Street Journal article that Musk saw deleted the initial post sharing the headline and the rumored sale of Tesla’s China business.
This has never even come up in a discussion ever.
Absurdly fake news.
-People should assume news is fake until proven otherwise.
— Elon Musk (@elonmusk) July 31, 2026
The report seemed absolutely and unequivocally false to begin with; Tesla’s business in China is among the most important pieces of the company’s business. Not only does the factory supply vehicles for the domestic market, but also for various other markets in Asia and Europe.
China is also one of the largest automotive markets in the world, and Tesla has performed well there despite the robust competition.
The speculation regarding a Tesla and SpaceX merger has started to gain steam this year as the space exploration company went public just a month ago. There has been speculation that Musk will bridge all of his companies under one “umbrella company,” and analysts believe this could happen before the end of the decade.
The Tesla and SpaceX merger everyone is talking about is quietly building
-
This is the latest iteration of Musk’s very evident war on mainstream media. Reports regarding any of Musk’s companies are quick to get the dreaded “false” or “fake news” response from the CEO when they are unfounded.
Elon Musk
Tesla AI boss reveals how big Optimus is going to get
Tesla’s Optimus chief corrected himself on X, confirming a staggering 10 million robot production target.
![Tesla Optimus Gen 3 [Credit: Tesla]](https://www.teslarati.com/wp-content/uploads/2026/03/tesla-optimus-gen3-diner.jpg)
Tesla’s Optimus program has a new number attached to it, after Ashok Elluswamy, the executive who has run the humanoid robot program since June 2025, posted a three word correction on X Thursday, “Correction, 10 million robots.”
The line clarifies the long term annual capacity Tesla is building toward its planned second Optimus production line at Gigafactory Texas, a figure Musk has cited repeatedly since last year’s shareholder meeting.
The scale is worth noting, because ten million robots a year would mean Tesla building more units annually than most countries sell in new cars. Tesla has framed this as a second line, not the first. The buildout is happening in two phases: a roughly one million unit per year line inside Tesla’s Fremont factory, installed on the floor space vacated when Model S and Model X production ended earlier this year, and a much larger dedicated facility under construction at Giga Texas that broke ground on its first steel structure in May. That Texas facility is the one Elluswamy’s correction refers to, and is expected to reach volume production sometime in 2027.
Correction, 10 million robots https://t.co/0z4nyQNTzp
— Ashok Elluswamy (@aelluswamy) July 30, 2026
Tesla Optimus project fires up as Musk sees production line progress
Elluswamy took over Optimus from Milan Kovac last summer and has spent the months since talking up the program’s trajectory. Elon Musk has also floated the ten million figure at Tesla’s 2025 shareholder meeting.
-
Ending Model S and Model X production to make room for the first Optimus line was one of the more consequential manufacturing decisions in the company’s recent history, retiring two flagship vehicles in favor of a robot that has yet to enter mass production. Musk has previously estimated per unit production costs at $20,000 to $25,000 once Tesla reaches a million units a year, though he hasn’t said what that cost looks like at ten times the volume.
Tesla Model Y L’s new features flexed at unveiling event at Diner
Tesla CEO Elon Musk denies ridiculous Gigafactory Shanghai rumor
![Tesla Optimus Gen 3 [Credit: Tesla]](https://www.teslarati.com/wp-content/uploads/2026/03/tesla-optimus-gen3-diner-80x80.jpg)
