Dangerous robots, dangers to privacy, and Frankenstein's Lego Game Boy

iRobot’s founder wouldn’t touch today’s robots with a 10-foot pole

Rodney Brooks said the billions being poured into developing bipedal humanoids is nothing more than an"expensive fantasy." His argument that making robots in our image is fundamentally unsafe, since if bipedal bots are made larger, they’re just bigger walking risks (imagine a 10-foot metal-and-polymer giant slipping and crashing into you), and the current energy systems used to power them can result in erratic behaviours, making them dangerous to be anywhere near them.

More importantly, they aren’t dextrous. The current training models employed by companies like Tesla and Figure rely purely on visual data, whereas humans learn from a plethora of multisensory information. Additionally, unless we can put as many sensors in a robot’s fingertips as humans have nerve endings, they will never be able to recreate our level of dexterity. 

So we need to take another direction with how we want to build bots, like how Wall-E is a small box on wheels. Rodney himself argues that future robots should not look like us, rather they should have wheels (like Wall-E!), multiple arms (not like Wall-E), and specialised sensors that don’t act like our eyes do (kinda like Wall-E?).

We should move towards Wall-E in the future.

Meta: “Yes, we will use your data. No, you can’t say no.”

I love it when large companies keep redrawing the lines on user privacy. 

Meta announced that on December 16th, what you say to the AI will be used to “personalise content and ad recommendations.” And the worst part? You can’t opt out. 

If this sounds like an excuse to push the boundaries of data usage and privacy, then yes, it is. 

Meta has said that they won’t use overly sensitive information for ad recommendations, like religious beliefs or sexual orientation; however, there have been many reports showing that not only do people overshare with AI, but there have been instances with ChatGPT and Meta where the AI posts private chats to public feeds or have them appear in Google search results.

And this isn’t just for the Meta app; this applies to any Meta product: smart glasses can record your surroundings, your voice, and use it to push any product it thinks is relevant to you.

But we all knew that they’re always listening and recording. For them to come outright and say it? That’s a new level of brazenness that should concern everyone.

Giving life to Lego (Game Boys)

The Lego Game Boy launched on Wednesday, and when X user @natalie_thenerd saw it, a flip switched inside her head like a sleeper agent getting activated.

Less than a day later, she’d managed to turn the Lego Game Boy into an actual, mostly functional Game Boy. Using an opening behind the screen for toy cartridges, she designed a custom printed circuit board using a pocket CPU complete with its own RAM. This means that she’s not simply emulating games on the Lego Game Boy, she’s using the original chips. Oh, and the entire circuit board is smaller than a Dot-Matrix Game Boy cartridge.

You’ll note that I said “mostly functional”, and that’s because the buttons, while you can press them, don’t register as inputs. Also, there isn’t any footage of her actually playing a game, but with what she’s created so far, I don’t think it’ll be long until we get a fully functional version.