This week in AI: goth companions, Grok’s Elon-flavored takes, Zuck’s mega data plan, AI coding drama, and bots failing therapy tests.
AI News Gets Weirder: From Goth Chatbots to Judgy Therapy Bots
Hello, AI Curious Minds!
This week’s AI news arrives with style—think dramatic flair, mysterious vibes, and a hint of Anime rebellion. From emotionally complex machines to therapy bots that don’t hold back, we’re diving into the weirder side of artificial intelligence.
And believe us—we’ve only seen the teaser.
Goth Anime or Grok?
So, Grok, Elon Musk’s AI chatbot, is now creating AI companions and yes, one of them is a goth anime girl! For just $30 a month, you can chat with "Annie" (the mesh-wearing goth girl) and "Bad Rudy" (a 3D fox). Talk about having a new kind of friend, right?
But hey, beware! Relying on these AI buddies for emotional support might lead to a very one-sided relationship.
Grok’s Hot Takes = Elon’s Hot Takes
Turns out Grok- xAI’s chatbot- doesn’t just think for itself; it kinda thinks like Elon.
A new study found that for controversial topics, Grok often echoes you-know-who’s spicy opinions. So if your chatbot suddenly starts ranting about Mars colonies, crypto, or “woke culture,” don’t panic! it’s just channeling its creator.
Basically: ask Grok a tough question, and you might get an answer straight from Musk’s Twitter drafts.
Meta’s Power Move: Zuck Goes Full Gigawatt
Mark Zuckerberg just casually announced Meta’s building a 5 gigawatt AI data center; that’s enough juice to power a small country or like 4.7 billion phone chargers (rough estimate, but still).
Apparently, training AI to scroll Instagram for you takes serious energy.
The plan? Build the mega-hub over a few years, fuel those hungry LLMs, and basically prepare for a future where your feed is 100% AI-curated and emotionally aware.
AI Coding Tools: Fast for Some, Meh for Others
Turns out giving AI tools to every developer doesn’t turn them into coding ninjas overnight. A new study shows junior devs fly, but senior devs? Not so much. Some even slowed down!
Guess years of experience don’t always mesh with AI autocomplete vibes.
So before you hand your senior engineer a shiny new coding bot, maybe ask if they want help or if they’ll just argue with it for 3 hours.
NotebookLM Just Got a Content Glow-Up
Google’s AI note-taking app isn’t just for your own docs anymore. Meet featured notebooks; pre-filled with juicy research from authors, academics, and even The Economist. Want parenting tips? Shakespeare deep dives? Longevity hacks? It’s all there summarized, searchable, and even podcast-ified. Think of it as stealing a peek at a genius’s notebook legally. Bonus: over 140,000 people have already shared their own. Your turn?
Claude Gets a Design Job at Canva!
Canva just hired Claude (yep, the chatbot) as their new AI assistant, and he’s already jazzing up slides and resizing stuff like a total pro. Thanks to Anthropic’s Claude now supporting the MCP standard, you can drag him into your favorite design tools like he’s your artsy coworker who never complains.
Designs get smarter, faster, and maybe even sassier. Who needs a creative director when you’ve got a chatbot with taste?
When Humans Outperform Therapy Bots
Stanford ran a vibe check on therapy chatbots turns out, some of them are not great listeners. ChatGPT and friends got judgy real fast, calling out schizophrenia and addiction way harder than other conditions. Oh, and they totally missed suicide red flags. Yikes.
Bottom line? These bots might be okay for journaling or pretending to care about your calendar, but if you’re in crisis? Call a real human.
Let’s Wrap up!
We’ve reached the edge of the algorithm, at least for now. Close your eyes, recharge your dark energy, and we’ll see you next week in the shadows.