Digital Digest #64 - Perplexity's rollercoaster; AI agents; deepfake concerns for models; Bluesky's growth

Hi everyone,

Welcome to this week's Digital Digest!

In this week's edition:

• Autonomous AI bursts onto the scene
• Deepfakes cause headaches for the models whose faces are being used by AI avatar platforms
• Bluesky boasts surging user numbers… but does anyone care?

But first, a look at a rollercoaster week for popular AI search platform Perplexity.

1. A perplexing week for Perplexity

Popular AI search tool Perplexity has had a rollercoaster week. In the space of a few days, Perplexity faced a lawsuit from several top news publishers (to go on top of the existing NYT suit), began talks over a targeted $1bn funding round, and said it now serves 100M search queries a week. (TechCrunch, eMarketer, Wall Street Journal - gift link, TechCrunch)

• Why it matters: Perplexity has rapidly become a go-to tool for millions of people (myself included) when looking for quick answers to questions, backed up by sources. However, it is increasingly being dogged by controversy as it runs into conflicts with news publishers who have concerns with it (and other platforms) scraping their content.

Also in search news:

• Marketers are not won over by new AI-powered search tools (Digiday)
• ‘The way we interact with search boxes has changed’: What GenAI and RAG mean for retailers (eConsultancy)
• Redditors are trying to poison Google to keep tourists out of the good restaurants (Ars Technica, Gizmodo)

2. Autonomous AI bursts onto the scene

Major AI announcements this week from Microsoft and Anthropic, both announcing new tools that allow generative AI assistants to handle tasks semi-autonomously. (Axios, Microsoft, Mashable, Anthropic, Anthropic)

• Why it matters: Hype around "AI agents" has been building throughout this year but reached a fever pitch in the last couple of weeks. These tools speak to a future where generative AI moves beyond chatbots and into even more productive functions for consumers and businesses. For now, they're in the early stages - Anthropic notes of its current release, "it is still?experimental—at times cumbersome and error-prone."

Also in generative AI news:

• How communicators are integrating AI into their work (Axios)
• OpenAI reportedly plans to release its Orion AI model by December (TechCrunch)
• Anthropic also introduced an analysis tool in Claude.ai that enables analysis and visualization of data (Anthropic)

3. AI deepfakes cause concerns for avatar models

Models whose likenesses have been used in AI avatar creation platforms are finding themselves used in AI propaganda and disinformation campaigns. (The Guardian)

• Why it matters: After some early examples of AI avatars being used for nefarious purposes, platforms rapidly put in place good-faith moderation policies and screening of videos for things like fake newscasts, profanity and other harmful content - but bad actors are still finding ways around them.

Also in misinformation news:

• AI sludge local news site Hoodline falsely accuses San Mateo DA of murder (TechDirt)

4. Bluesky gains users but remains anaemic

Perplexity wasn't the only platform with a busy week - Bluesky announced it gained 2M users after X's latest user-alienating announcement, teased a new? creator payments feature and announced a new round of funding. (Social Media Today,?Gizmodo,?TechCrunch)

• Why it matters: Unfortunately for Bluesky, user numbers remain insignificant in comparison to X and Threads which boast users in the hundreds of millions (Threads recently crossed 200M). I also doubt the average person on the street has ever heard of it -- unless something changes, it seems destined to be just another fringe social media platform.

Also in social media news:

• Elon Musk's X further squeezes developers with apparent new API fees (Mashable)

Have a great week!

Dave