Tuesday, May 21, 2024

Welcome to TechCrunch AM! This morning, we’ve got notes on data-labeling firm Scale AI’s huge fundraise, EU leading the crowd on AI regulation, how the NYC-Dublin portal has opened with new protections, a look at the deepfakes being used to manipulate India’s elections, and how Microsoft plans to bring Windows into the current century. — Rebecca

TechCrunch Top 3

Image Credits: David Paul Morris/Bloomberg via Getty

1. Scale AI raises $1B, doubles valuation: Data-labeling company Scale AI has raised a $1 billion Series F, doubling its valuation to $13.8 billion, in no surprise to anyone. Data is the lifeblood of AI, and any company managing and processing it is cashing in today. Read More

2. EU gets nod to set up AI regulations: The EU Council has given final approval to set up the bloc’s flagship, risk-based regulations for AI. The law bans certain “unacceptable risk” cases outright, like cognitive behavioral manipulation, and defines a set of “high risk” uses like biometrics or facial recognition in domains like education and employment. Expect all the AI companies to now redirect their lobbying budgets to influence how this new law is laid out. Read More

3. ‘Tinder Swindler’ is bad for business: That’s why a group of tech companies, including Match Group, Meta and Coinbase, are banding together to take on online fraud. Tech Against Scams, as it’s called, will work together to fight back against tools used by scammers and to better educate the public about financial scams. Read More

A MESSAGE FROM DELINEA

Rising Challenges in Cyber Insurance: What Companies Must Know

A comprehensive survey sheds light on the critical aspects of cyber insurance, revealing more than just a 'check-the-box' necessity for organizations.

Learn More

Morning Must Reads

Image Credits: Spencer Platt / Getty Images

The NYC-Dublin portal reopens: The NYC-Dublin video portal closed shortly after its launch after some people behaved inappropriately — as expected. The portal has now reopened under new conditions: There’s a fence around it to discourage people from going right up to it and flashing their nethers, a few people are guiding the experience, and the hours are limited. Read More

VinFast gets pulled up: U.S. safety regulators are investigating a crash involving an electric VinFast VF8 SUV that killed a family of four in April. The Vietnamese automaker began shipping its SUVs to the U.S. in early 2023, and had issued a recall over problems with the dashboard screen going blank. Read More

There’s work in robotics: There’s never been a more exciting time to be in robotics, whether you’re into automating the production or building humanoid ‘bots. Want to be a part of the future? TC’s Brian Heater has compiled available jobs from 81 companies.Read More

Bringing Windows into the 21st century: Windows has become something of a meme for how little it has changed in the last decade. But now, Microsoft wants to bring Windows to 2024 by adding generative AI to it, and recently unveiled a slew of GenAI features for the operating system at its Build developer conference. Yes, that will definitely make Windows less of a meme. Read More

ETF Partners gets big bucks for climate: The venture firm has raised an oversubscribed €285 million fund that will invest in climate tech startups. But instead of focusing on tech with long runways, like carbon capture or making electric planes, ETF is more interested in climate software companies that can move the needle sooner than later. Read More

Aurora reveals self-driving Volvo truck: How ready are you to see a driverless truck chugging along next to you on the highway? Aurora Innovation is working on a new self-driving truck that’s being manufactured by Volvo and could be testing on public highways as early as this summer. It’s an important step towards Aurora’s goal to launch 20 fully driverless big-rigs by the end of this year. Read More

Can tech solve the teacher shortage crisis? We should probably just pay teachers more. But until that happens, London-based Zen Educate has raised $37 million to replace traditional third-party recruitment agencies with its marketplace that connects schools with teachers. Read More

Fintech for insurance gets $$$: Whoever thinks fintech startups are dying needs to read the news more often. U.K. fintech Vitesse has raised a whopping $93 million to build out its all-in-one treasury and payment management platform that can streamline international payments, among other fintech-y services, for insurers. Read More

Around the Web

The Indian canary in the coal mine: Meta approved a series of AI-manipulated political advertisements – AKA deepfakes – during India’s election that spread disinformation and incited religious violence, The Guardian reports. I won’t go into the horrifying ads that spread across the platform in India, but the rest of the world should be concerned, if not outright terrified — elections are either happening or about to kick off soon everywhere. Read More

Google cuts the DOJ a mystery check: Google has preemptively paid damages to the U.S. government, Reuters reports. Google is nearing the end of its antitrust trial, which alleges the company has been (surprise, surprise) monopolistic with its ad tech. This unusual move seems to be aimed at avoiding a jury trial. Read More

OpenAI sneaks up on Nvidia: Nvidia is the Big Dog in chipmaking right now, but the company’s rivals and biggest customers are taking a swing at its hold on software. OpenAI is leading that fight, apparently: Engineers at Meta, Microsoft and Google, which have all spent billions of dollars on Nvidia chips, are contributing to OpenAI’s competing software platform, Triton, reports the Financial Times. Read More

A MESSAGE FROM ESPER

Overcoming Challenges with Operationalizing AI
at the Edge

Bridging the gap from an AI pilot to real-world impact is challenging, but not if you have the right tools to optimize deployment, reduce costs, and update with fluidity. It’s not impossible — it’s reality. Let’s break it down.

Image Credits: Amy Sussman / Getty Images

OpenAI was actually trying to imitate ‘Her’: OpenAI said yesterday it would remove the Sky voice on ChatGPT that sounded eerily like Scarlett Johansson in the movie "Her." Well, Johansson has now said OpenAI actually approached her about it! The outrage around the human-like voices OpenAI is choosing to deploy is important, because they’re not necessary for the tech to be useful. It’s just meant to make you feel an emotional attachment to the AI so you use it more. Read More