This week in AI: OpenAI and publishers are partners of convenience

Keeping up with an industry as fast-paced as artificial intelligence is a tall order. So, until an AI can do it for you, here’s a helpful roundup of recent stories in the world of machine learning, along with notable research and experiments we didn’t cover on their own.

By the way, TechCrunch plans to launch an AI newsletter soon. Stay tuned. In the meantime, we’re increasing the pace of our semi-regular AI column, previously twice a month (or so), to weekly — so be on the lookout for more releases.

This week in AI, OpenAI was announced that it struck a deal with News Corp, the new publishing giant, to train artificial intelligence models developed by OpenAI on articles from News Corp brands, including The Wall Street Journal, Financial Times and MarketWatch. The agreement, which the companies describe as “multi-year” and “historic”, also gives OpenAI the right to display News Corp mastheads on apps such as ChatGPT in response to certain questions — possibly in cases where the answers are partially or in their entirety by News Corp. Publications.

Sounds like a win for both parties, no? News Corp gets an infusion of cash for its content — more than $250 million, According to reports — at a time when the outlook for the media industry is even more gloomy than usual. (Generative AI didn’t help mattersthreatening to significantly reduce the citation traffic of publications.) Meanwhile, OpenAI, which has been battling copyright holders on a number of fronts over fair use disputes, has a less costly legal battle to worry about.

But the devil is in the details. Note that the News Corp agreement has an expiration date — as do all OpenAI content licensing agreements.

This in itself is not bad faith on OpenAI’s part. Perpetual licensing is rare in the media, given the incentive of all parties involved to keep the door open to renegotiating the deal. However, this is a little suspicious in light of OpenAI CEO Sam Altman’s recent comments about the declining importance of AI model training data.

In an appearance on the “All-In” podcast, Altman he said that he “certainly [doesn’t] think there will be an arms race for [training] data” because “when models get smart enough, at some point, it shouldn’t be about more data — at least not for training.” Elsewhere, he he said MIT Technology Review’s James O’Donnell that he is “optimistic” that OpenAI — and/or the broader AI industry — will “find a way out of [needing] more and more training data”.

The models aren’t that “smart” yet, which leads OpenAI to reportedly experimenting with synthetic training data and scan web losses — and YouTube — for biological sources. But suppose they do one day no a lot of additional data is needed to improve by leaps and bounds. Where does this leave publishers, especially when OpenAI has ripped through their entire archives?

My point is that publishers – and the other content owners that OpenAI has partnered with – appear to be short-term partners of convenience, not much more. Through licensing agreements, OpenAI effectively neutralizes a legal threat — at least until the courts determine how fair use applies in the context of AI education — and can celebrate a PR victory. Publishers receive the necessary capital. And work on AI that can seriously hurt these publishers continues.

Here are some other notable AI stories from the past few days:

• Spotify’s DJ AI: Spotify’s addition of its AI DJ feature, which introduces personalized song selections to users, was the company’s first step into an AI future. Now, Spotify is developing an alternative version of this DJ that will speak Spanish, writes Sarah.
• Meta AI Board: Meta announced Wednesday the creation of a AI Advisory Board. However, there is one big problem: it only has white men on it. This is a bit tone-deaf, given that marginalized groups are the ones most likely to suffer the consequences of AI technology’s shortcomings.
• FCC Proposes AI Disclosures: The Federal Communications Commission (FCC) has imposed a requirement to disclose AI-generated content in political ads — but not ban it. Devin has the full story.
• Answer calls by voice: Truecaller, the popular caller ID service, will soon let customers use its AI assistant to answer their phone calls. mine voice, thanks to a new partnership with Microsoft.
• Humane is considering selling: Humane, the company behind the much-hyped Ai Pin that launched to less than glowing reviews last month, is on the hunt for a buyer. The company is reportedly valued at between $750 billion and $1 billion, and the sale process is in the early stages.
• TikTok Turns to Genetic AI: TikTok is the latest tech company to incorporate genetic artificial intelligence into its ads, as the company announced Tuesday that it is launching a new TikTok Symphony AI suite for brands. The tools will help marketers write scripts, produce videos and improve their current advertising assets, reports Aisha.
• Seoul AI Summit: At an AI security summit in Seoul, South Korea, government officials and AI industry executives agreed to implement rudimentary security measures in the rapidly evolving field and create an international security research network.
• Microsoft’s AI computers: In two keynotes during its annual Build developer conference this week, Microsoft unveiled a new line of Windows machines (and Surface laptops) called Copilot+ PCs, as well as AI-powered productivity features like Recall, which helps users to find apps, files, and other content they’ve seen before.
• OpenAI’s voice disaster: OpenAI removes one of the voices in ChatGPT’s text-to-speech feature. Users found the voice, named Sky, to be uncannily similar to Scarlett Johansson (who has played AI characters in the past) — and Johansson herself released a statement saying she had hired legal counsel to inquire about Sky’s voice and to get exact details about how it was developed.
• UK Autonomous Driving Law: UK self-driving car regulations are now official after receiving royal assent, the final rubber stamp that any legislation must pass before it is enshrined in law.

More machine learning

Some interesting pieces of AI research for you this week. Prolific University of Washington researcher Shyan Gollakota strikes again with a pair of noise-cancelling headphones you can ask for block out everything but the person you would like to hear from. While wearing the headset, you press a button while looking at the person and it samples the voice coming from that particular direction, using it to feed an acoustic blocking engine so that background noise and other voices are filtered out.