Three Tech News Stories to Keep Your Eye On 

Internet Archive, New York Post, and Anthropic are all in the news.

Posted on October 28, 2024 in Blog Posts

Authors:

Daniel Pfeiffer
A librarian keeping up to date on tech news about AI and the Internet Archive

With an upcoming presidential election in the US, many important news stories have been lost to headlines and opinion pieces. October began with OpenAI, the creator of ChatGPT, announcing a $157 billion valuation after completing its most recent fundraising deal. But other developments this past month are easier to miss, so we wanted to highlight them before they’re lost in the digital churn. 

👾 The Internet Archive’s problems compound. 

The Internet Archive has had a rough few months. After failing its legal appeal against publishing houses, which has resulted in the removal of more than 500,000 books from its online collection, it suffered a massive hack in early October. A hacker stole identifying information—including emails, passwords, and other internal data—for more than 31 million users. The hacker left a message in the website’s code, taunting Internet Archive for its poor security practices. The Internet Archive responded by shutting down several of its services as it strengthened its security. Shortly after announcing last week that it was back online, another hack breached its customer service platform. Though the Wayback Machine remained online, most of the Internet Archive’s services were down until last Thursday.

🌟 Subscribe to the LibTech Insights newsletter for weekly roundups and bonus content, including: 

👩‍⚖️ Dow Jones and New York Post hit Perplexity with a lawsuit. 

Perplexity has received some positive attention from critics (including our own!) as an alternative to ChatGPT. A well-known problem with generative AI models is that they’re prone to making up information—a phenomenon known as “hallucination.” Perplexity tries to overcome this problem by having its AI model give citations for its claims, allowing users to fact check its responses. Additionally, its open web sources allow for more up-to-date content. ChatGPT, in contrast, uses a limited dataset and currently contains only knowledge up to April 2023. 

In an unsurprising move, Rupert Murdoch’s Dow Jones and New York Post sued Perplexity. The lawsuit alleges that Perplexity plunders the publishers’ content and repackages it through its AI responses. I said this move was “unsurprising,” because copyright lawsuits against generative AI companies are quickly becoming normal. But this case is interesting for showing the legal entanglements that arise when trying to solve some of generative AI’s most foundational problems.  

🖥️ Anthropic’s Claude can now control your desktop. 

Anthropic’s generative AI model Claude has largely flown under the radar. Its value-add among AI models has been hard to locate. This may change with its newly launched beta feature: computer use. This will allow Claude to use your computer as you would—“by looking at a screen, moving a cursor, clicking buttons, and typing text.” For instance, if you ask it to send the 2024 fiscal year report to René, it will open Outlook, create an email to René, find the report in your Documents folder, attach it, and send the email. See it in action. 

It is an open question, I think, whether users are ready to let an AI control their computers, particularly given the cybersecurity risks. Nonetheless, this feature is a notable advance in generative AI technology, and its development should be on everyone’s radars. 

🔥 Sign up for LibTech Insights (LTI) new post notifications and updates.

✍️ Interested in contributing to LTI? Send an email to Daniel P. at Choice with your topic idea.

Related Posts