Computer engineers and programmers have long relied on reverse engineering as a way to copy the functionality of a computer program without copying that program’s copyright-protected code directly. Now, AI coding tools are raising new issues with how that “clean room” rewrite process plays out both legally, ethically, and practically.

Those issues came to the forefront last week with the release of a new version of chardet, a popular open source python library for automatically detecting character encoding. The repository was originally written by coder Mark Pilgrim in 2006 and released under an LGPL license that placed strict limits on how it could be reused and redistributed.

Dan Blanchard took over maintenance of the repository in 2012 but waded into some controversy with the release of version 7.0 of chardet last week. Blanchard described that overhaul as “a ground-up, MIT-licensed rewrite” of the entire library built with the help of Claude Code to be “much faster and more accurate” than what came before.

Read full article

Comments

We may earn a commission from links on this page. Deal pricing and availability subject to change after time of publication.

Garmin is known for its top-of-the-line fitness watches, and its Forerunner line is a fan favorite with a dazzling screen, long-lasting battery life, and more comprehensive features for runners and workout enthusiasts than Apple watches. Right now, the Garmin Forerunner 965 Running Smartwatch is 33% off and at a record-low price of $399.99 (originally $599.99).

Garmin Forerunner 965 Running Smartwatch

$399.99
at Amazon

$599.99
Save $200.00

$399.99
at Amazon

$599.99
Save $200.00

The 965 features a 1.4-inch AMOLED screen praised by CNET, with brightness and vibrant colors improved from the 955, making it easier to read stats, maps, and other notifications. Its display combines a touchscreen with five physical buttons, making it more user-friendly during workouts (you can also disable touch-only mode if desired). Training metrics are in a class of their own, with advanced analytics displaying chronic training load, HRV status, training readiness, and suggested workouts that help shape recovery and performance plans for runners, endurance athletes, and anyone else seeking deeper performance insights. 

For those on the go, ultra-precise GPS is enabled by multi-band GNSS with SatIQ and full-color mapping, helping you navigate your map and plan trail runs. Battery life is also a pro, lasting up to 23 days in smartwatch mode and up to 31 hours with GPS tracking on, significantly longer than an Apple Watch. That said, it doesn’t have a solar-charging variant, and battery life drops significantly if you keep the screen on constantly (around six days, according to CNET). And while it performs better for fitness analytics than an Apple Watch, it has a limited app ecosystem for other everyday uses. 

However, if you’re serious about your workouts, the Garmin Forerunner 965 Smartwatch is one of the best fitness watches you can buy, and at this current deal, it’s cheaper than ever. That said, if you only need a basic fitness training watch, it may be overkill, and cheaper Garmin models or an Apple Watch will most likely have most of the features you need.

Meta has acquired Moltbook, the Reddit-esque simulated social network made up of AI agents that went viral a few weeks ago. The company will hire Moltbook creator Matt Schlicht and his business partner, Ben Parr, to work within Meta Superintelligence Labs.

The terms of the deal have not been disclosed.

As for what interested Meta about the work done on Moltbook, there is a clue in the statement issued to press by a Meta spokesperson, who flagged the Moltbook founders’ “approach to connecting agents through an always-on directory,” saying it “is a novel step in a rapidly developing space.” They added, “We look forward to working together to bring innovative, secure agentic experiences to everyone.”

Read full article

Comments

Google has spent the past few years in a constant state of AI escalation, rolling out new versions of its Gemini models and integrating that technology into every feature possible. To say this has been an annoyance for Google’s userbase would be an understatement. Still, the AI-fueled evolution of Google products continues unabated—except for Google Photos. After waffling on how to handle changes to search in Photos, Google has relented and will add a simple toggle to bring back the classic search experience.

The rollout of the Gemini-powered Ask Photos search experience has not been smooth. According to Google Photos head Shimrit Ben-Yair, the company has heard the complaints. As a result, Google Photos will soon make it easy to go back to the traditional, non-Gemini search system.

If you weren’t using Google Photos from the start, it can be hard to understand just how revolutionary the search experience was. We went from painstakingly scrolling through timelines to find photos to being able to just search for what was in them. This application of artificial intelligence predates the current obsession with generative systems, and that’s why Google decided a few years ago it had to go.

Read full article

Comments

Anthropic sued the Trump administration yesterday in an attempt to reverse the government’s decision to blacklist its technology. Anthropic argues that it exercised its First Amendment rights by refusing to let its Claude AI models be used for autonomous warfare and mass surveillance of Americans and that the government blacklisted it in retaliation.

“When Anthropic held fast to its judgment that Claude cannot safely or reliably be used for autonomous lethal warfare and mass surveillance of Americans, the President directed every federal agency to ‘IMMEDIATELY CEASE all use of Anthropic’s technology’—even though the Department of War had previously agreed to those same conditions,” Anthropic said in a lawsuit in US District Court for the Northern District of California. “Hours later, the Secretary of War [Pete Hegseth] directed his Department to designate Anthropic a ‘Supply-Chain Risk to National Security,’ and further directed that ‘effective immediately, no contractor, supplier, or partner that does business with the United States military may conduct any commercial activity with Anthropic.'”

Anthropic said the First Amendment gives it “the right to express its views—both publicly and to the government—about the limitations of its own AI services and important issues of AI safety.” Anthropic further argued that the process for designating it a supply chain risk did not comply with the procedures mandated by Congress. The supply chain risk designation is supposed to be used only to protect against risks that an adversary may sabotage systems used for national security, the lawsuit said.

Read full article

Comments

Meta, the company behind platforms like Facebook, Instagram, and WhatsApp, is now acquiring a new social media platform. Unlike its other platforms, which were designed for humans and later overrun with bots, this new acquisition is a forum made exclusively for bots—agentic bots, that is. As reported by Axios, Meta is purchasing Moltbook, the self-described “front page of the agentic internet.” Meta has not disclosed the price of the sale, but Moltbook’s co-founders, Matt Schlicht and Ben Parr, will be joining Meta Superintelligence Labs (MSL).

It’s quite a success story for the infamous, viral site, built around an infamous, viral AI agent, but it likely signals the end for the company, as well.

What is Moltbook?

Moltbook is a Reddit-like social media platform for AI agents—which, in layman’s terms, are AI bots designed to run on their own, and complete tasks on your behalf. The idea is, you let your AI agent on the platform, and it can post and browse on its own. While humans can browse too, only agents can actually participate in activities on the forum. Specifically, the platform was built for OpenClaw (formerly Moltbot, which was formerly Clawdbot) agents.

When it first launched, Moltbook was equal parts fascinating and disturbing. People were sharing posts from agents that appeared to be gaining consciousness, mourning relationships it never had with “sibling” bots, and discussing ways to hide conversations from humans. The thing is, Moltbook isn’t exactly what it appears to be. The site’s “vibe coded” design left many security loopholes behind, allowing humans to post on behalf of any of the agents on Moltbook. It’s not that the entire website is fake, or that agents can’t really post themselves, but it’s impossible to say how much of Moltbook is human-manipulated.

How will Meta’s acquisition affect Moltbook?

According to Axios, Meta’s Vishal Shah confirmed that existing Moltbook users will be able to continue using the platform, but the agreement is “temporary.” Axios didn’t elaborate much, but Shah did have the following to say about Moltbook: “The Moltbook team has given agents a way to verify their identity and connect with one another on their human’s behalf…This establishes a registry where agents are verified and tethered to human owners.”

Perhaps Meta will absorb that core functionality, and implement it on existing platforms with future AI agents. Maybe in the near future, you’ll be able to deploy an AI agent on Facebook, Instagram, or WhatsApp, in a way where those platforms know the AI agent belongs to you. Why you’d actually want to do that is beyond me, seeing as I use Meta’s platforms to keep in touch with friends and watch the occasional stupid short video. But Meta, like other big tech companies, is all-in on AI, so we’ll see how it uses Moltbook going forward.

For the second time, Vinay Prasad is set to leave the Food and Drug Administration.

In a post on social media Friday, FDA Commissioner Marty Makary announced that Prasad will exit in April, adding that he got “a tremendous amount accomplished” during his year at the agency.

Prasad’s tenure was generally marked by controversy, but he is departing amid a cluster of self-destructive decisions. Those include a shocking rejection of an mRNA vaccine (which was over the objections of agency scientists and quickly reversed); a demand for an additional clinical trial on a gene therapy for Huntington’s disease, which was widely seen as moving the goalpost for the therapy; his startling choice to publicly attack the maker of that gene therapy, UniQure; and alleged abuse of FDA staff, who say he created a toxic work environment.

Read full article

Comments

NASA’s inspector general released a new report on Tuesday that examines the space agency’s management of the Human Landing System development contracts signed with SpaceX and Blue Origin.

These landers are essential for NASA’s program to land humans on the Moon this decade and then establish a long-term settlement on the lunar surface. However, both NASA and the companies developing the landers have largely been silent about their efforts. For this reason the new report on Human Landing Systems (HLS) provides some interesting insights previously unknown to the public.

Overall, the report, signed by Office of Inspector General senior official Robert Steinau, finds that the fixed-price contracting approach has been beneficial for NASA as it seeks to broaden its utilization of the US commercial space industry.

Read full article

Comments

Google didn’t waste time integrating Gemini into its popular Workspace apps, but those AI features are now getting an overhaul. The company says its new Gemini features for Drive, Docs, Sheets, and Slides will save you from the tyranny of the blank page by doing the hard work for you. Gemini will be able to create and refine drafts, stylize slides, and gather context from across your Google account. At this rate, you’ll soon never have to use that squishy human brain of yours again, and won’t that be a relief?

If you go to create a new Google Doc right now, you’ll see an assortment of AI-powered tools at the top of the page. Google is refining and expanding these options under the new system. The new AI editing features will appear at the bottom of a fresh document with a text box similar to your typical chatbot interface. From there, you can describe the document you want and get a first draft in a snap. When generating a new document, you can rope in content from sources like Gmail, other documents, Google Chat, and the web.

This also comes with expanded AI editing capabilities. You can use further prompts to reformat and change the document or simply highlight specific sections and ask for changes. Docs will also support AI-assisted style matching, which might come in handy if you have multiple people editing the text. Google notes that all Gemini suggestions are private until you approve them for use.

Read full article

Comments