Tesla’s value predicted to boost from new Dojo self-driving supercomputer

Analysts predict Tesla’s new ‘Dojo’ image-learning supercomputer could boost the electric car company’s market value by almost AU$1 trillion.

The prediction comes from financial firm Morgan Stanley [via Reuters ↗] – which catalysed a 10 per cent spike to Tesla’s stock price overnight, adding more than AU$100 billion to its market valuation, according to Markets Insider [↗].

2

Production of the Dojo supercomputer started in July and Tesla’s chief executive Elon Musk live-streamed a demonstration on X last month teasing the human-like driving behaviour of the new image-based learning model.

Tesla plans to spend more than AU$1.5 billion in the next year on Dojo and Morgan Stanley analysts predict it could open up revenue opportunities for the company beyond selling electric vehicles and energy products.

The company hovered around AU$400 per share last week and is currently trading at about AU$426.

Despite selling globally-popular electric vehicles, Tesla’s profits have reportedly mostly come from selling carbon offset credits to other automakers, its public Supercharging network, and energy business.

In April, Reuters [↗] revealed groups of Tesla employees internally shared invasive and private videos of recordings from owners’ cars.

They were allegedly part of the FSD Beta team, where data image labellers – who train the software to detect objects and signs – could access customer vehicle recordings.

FSD Beta has been available since late 2020 in North America – but has been under scrutiny from regulators as it may ‘act unsafe’.

The company has long claimed an ‘imminent’ full autonomous driving future, with so-called ‘robotaxis’ transporting passengers.

In Australia, new and existing Tesla electric vehicle owners have the option to purchase a ‘Full Self-Driving Capability’ software package for $10,100 – but only ‘promises’ FSD.