The Download: how Twitter is breaking, and YouTube’s TV experiment

This year we’ve seen a dizzying number of breakthroughs in generative AI, from AIs that can produce videos from just a few words to models that can generate audio based on snippets of a song.

Melissa Heikkila, MIT Technology Review’s senior AI reporter, stopped by Google’s new Manhattan offices last week, where the company announced a slew of advances in generative AI, including a system that combines its two text-to-video AI models, Phenaki and Imagen.

While they’re impressive pieces of AI research, it’s unclear how Google could monetize them. Melissa spoke with several top executives from some of the most prestigious AI labs around the world to learn more about the potential and limitations of these models. Here’s what they had to say.

Melissa’s story is from The Algorithm, our weekly AI newsletter covering everything you need to know about the industry’s movers and shakers. Sign up to receive it in your inbox every Monday.

I’ve combed the internet to find you today’s most fun/important/scary/fascinating stories about technology.

1 There’s no evidence that US voting machines have been tampered with
Humans tend to be the weakest link in the security chain. (New Yorker $)
Apps popular among immigrants are rife with political misinformation. (WP $)
The worst surge of misinformation could be yet to come. (NYT $)

2 Cop27’s Wi-Fi in Egypt is blocking human rights websites
Global rights groups are struggling to access their own sites. (The Guardian)
Greece will stop selling spyware following a series of accusations. (NYT $)

3 A German privacy activist is fighting Clearview AI over his face
Matthias Marx wants EU regulators to crack down on data scrapers. (Wired $)
A UK group has filed a similar complaint against PimEyes. (BBC)
The walls are closing in on Clearview AI. (MIT Technology Review)

5 Apple is China’s most profitable tech company
Its earnings outstrip native giants Alibaba and Tencent. (FT $)
But the relationship between the pair is growing increasingly fraught. (NYT $)

6 Inside the rise of the humanoid robot
They may be edging beyond the uncanny valley. (Economist $)

7 Self-driving cars may never actually self-drive
Which kind of defeats the entire point. (WSJ $)
The big new idea for making self-driving cars that can go anywhere. (MIT Technology Review)

8 How TikTok ate the world
Attempting to moderate videos is tough. TikTok’s explosive growth is making it harder. (The Atlantic $)

9 How psychedelics could play a role in end of life care
Some doctors argue they reduce anxiety and increase optimism in the face of death. (Slate $)
What do psychedelic drugs do to our brains? (MIT Technology Review)

April 2022

It was meant to be a temporary side job. Oskarina Fuentes Anaya joined Appen, an AI data-labeling platform. She was still in college and wanted to get a job in the oil industry.

But then, the economy crashed in Venezuela. Her side gig was now full-time; the temporary now the future. Fuentes is now a Colombian refugee and migrant, living in Colombia.

She is trapped at home due to a chronic illness she developed after she was denied access to medical care. She also has to work and earn a living through opaque algorithms.

Despite threats from Appen to retaliate against her, she chose to go on the record as a named source. She wants people to see her life as a key part of the global AI pipeline, but for her beneficiaries to also mistreat and make her invisible. She wants to see the people who do this work. Read the full story.