Welcome to SundAI, brought to you with a sprinkle of sarcasm and a side of dystopian dread.
Are you ready for some tech that will save us or doom us. This time, with less nerd speak, and more stuff you’ll want for yourself, and with a few gnarly comments from moi.
Let’s roll. YOLO!
And, as usual…. ze commercial:
If you like this article and you want to support me..
- Comment, or share the article; that will really help spread the word 🙌
- Connect with me on Linkedin 🙏
- Subscribe to TechTonic Shifts to get your daily dose of tech 📰
Dream Machine 1.5. Now with more realism and less of a soul
Luma Labs is back at it, and they are cranking out AI-generated dreams. Except this time, they are faster, and eerily lifelike. With Dream Machine 1.5, you can now create five seconds of video in two minutes. That’s right, two minutes to create content so that is so realistic, that it will make you question if your own memories were generated by an AI and we are all living in a simulation.
They have also added a Photon Image Model. No, it’s not from Star Trek, but it promises high-quality image generation that’s faster than your Wi-Fi can handle the traffic. Oh, and the interface got a lot better.
It’s available for everyone from TikTokers to enterprises. The pricing ranges from “broke creative” (that’s me, donate the cup of coffee, please, be the first! – see down below), to “corporate overlord” with too much money.
Luma Labs is cornering the market on AI storytelling!
Pickle avatars. A godsent. Because who needs to attend meetings anyway?
This week’s “Are we seriously doing this”? award goes to Pickle. Now, that’s a startup that lets you send a digital clone to your Zoom calls. For just $300 to $1,150 a year, you can ditch those soul-sucking video conferences while your avatar nods along on your behalf.
Hahahahahahaaaaa. Yeeeeeey for progress !
Got a boring strategy meeting? Let Pickle take over while you binge Netflix and sip cocktails on the beach. Just upload five minutes of your face, and voila, it’s got a a virtual you that looks vaguely lifelike will be “engaged” on Google Meet.
Watch the video down below to get an idea how it works, and what it looks like:
Hilarious!
There is always a catch though..
It’s Mac-only for now, so Windows users will need to keep their cameras on a little longer.
Pickle is betting big on a future where showing up is optional. But let’s be real, it’s only a matter of time before someone’s avatar gets caught “sleeping” in a virtual boardroom.
I’m loving it already!
Tax season just got a robotic update with Fondo AI
If you happen to love Anime, you might be acquainted with Fondo de Pandella. But this here Fondo, is not really so animated, cause it’s an AI-driven bookkeeping startup. And they are here to rescue small businesses from the accountant shortage with algorithms. They already have landed around 1,200 customers and secured $6M in annual recurring revenue. AND they’ve just raised $7M to expand their empire.
They have an AI agent that does some number crunching and it also delivers you financial insights like some sort of digital CFO.
Sounds great, yes? Unless, of course, the Fondo starts “suggesting” you funnel your savings into crypto scams!
Reid Hoffman vs. Elon Musk
In the latest episode of “Tech Bros Behaving Badly”, on the one side of the cage, you have Reid Hoffman and he called out ze Elon Musk for advising President-elect Trump on AI policy while running his own AI startup, xAI.
If you are asking yourself what Hoffman’s concern really was about, here yougo: Musk might turn U.S. policy into a buffet line for his own companies, and leaving the rest of us with nothing but crumbs. You can bet your arsch that he will just do that. And Musk, ever the disruptor, has already joined the “Department of Government Efficiency”.
Efficiency for whom, Elon?
Stay tuned as this saga unfolds, proving once again that the intersection of politics and AI is a huge trainwreck in the making, that we can’t stop watching.
Apple’s Siri just got a lot smarter, chattier, but still kind of meh
As you know, Apple’s a bit late to the AI game and they are playing catchup. Their latest move to become relevant in the AI game is by trying to supercharge Siri with Large Language Models. Translation: Siri’s about to get chattier and less useless.
Watch the video:
Their goal is to move beyond “Hey Siri, what’s the weather” to “Hey Siri, draft my resignation letter and send it to HR”. Ambitious, sure, but can Siri really compete with Alexa and Google Assistant when it still struggles to understand “play my Spotify playlist”?
Let’s hope Apple delivers, or Siri might remain the AI equivalent of a rotary phone in a smartphone world.
OpenAI gets creative: Rap God 2.0
OpenAI just upgraded GPT-4o with a creativity boost, and it’s spitting out Eminem-style bars about quantum mechanics that makes Slim Shady jealous.
Kyle Shannon pushed the model to write “The Quantum Cipher”, and that is a rap about quantum mechanics. And he made it complete with intricate rhyme schemes and machine-gun delivery. Yes, yes! And then he handed it over to Suno’s v4 AI music generator, which turned the lyrics into a banger.
Listen it here: https://x.com/kyleshannon/status/1859355131738734824
My take: Congrats, OpenAI, you’ve just made writers and songwriters even more obsolete. Suno’s audio quality is really catching up to professional tracks, and we’re not far from AI albums topping Spotify.
Human musicians, enjoy your gigs while you can.
Google vs. OpenAI and the battle for AI supremacy
In the never-ending cage match for AI dominance, we have on the one side OpenAI with their release of its GPT-4o update to reclaim the top spot from Google’s Gemini 1114. Google retaliated with Gemini 1121, which immediately snatched the crown back.
Here’s a pic that I know you’ll not try to decipher anyway:
My take: Benchmarks are fun and all, but until Gemini stops feeling like it’s been programmed by a committee, I’ll stick with GPT-4. The real winner will be the one that doesn’t melt my brain when I try to write a simple script.
Runway’s Expand Video Feature: AI Outpainting Goes Hollywood
Runway just launched a new creature feature called Expand Video, and that beast generates video content beyond the edges of your original clip. So you upload any video you got of you and your wife doing nothing important, and want to spice things up… just hit the “GO” button and now you have turned your lousy square TikTok with additional video, into a widescreen 4k masterpiece?
Done.
More Suno. With v4, AI music gets louder and sharper
Suno’s v4 update is here! And it is promising us cleaner audio, more dynamic song structures, and a remastering tool for old tracks, because it knows you still like your old 8-track. They have also added a creative lyrics model, ReMi, to help Suno flex on human songwriters.
My take: Suno is coming for the music industry, and they’re doing it quite seriously. You can expect lawsuits from record labels any minute now. In the meantime, enjoy the AI-generated tunes before they disappear behind copyright battles.
Elon Musk is ranked Diablo IV player worldwide
Elon Musk is now the #1 Diablo IV player globally.. I didn’t know he was that kind of a nerd. But thanks to a mere 100 hours a week grinding dungeons, he managed to get hold of the place. But maybe, just maybe it’s because he is secretly eye-balling game studios to buy them and turn them into anti-procedural AI-generated monsters.
Meanwhile, SpaceX caught a Starship booster mid-air, but let’s focus on what really matters: Elon smashing virtual demons. I think that Musk’s gaming grind puts pro streamers to shame. I’d love to see him stream a board meeting on Twitch, but I guess that wouldn’t rake in the same clout as Diablo IV’s leaderboards.
Perplexity’s shopping assistant
Perplexity launched a new shopping assistant. It comes complete with visual search and one-click checkout. Amazon who?? Retailers can pay (sorry, “provide data for free”) to boost their products in search results. Perplexity is literally owned by Jeff Bezos and is now officially Amazon’s little cousin since it took a bribe of 6 Billion to “update it’s model”. Perplexity doesn’t know how recommendations work yet, but it will happily take your money to figure it out. Expect ads, affiliate fees, and shady rankings in 3… 2… 1…
Black Forest Labs’ launched FLUX.1 tools. Photoshop on cocaine
FLUX.1 Tools just launched with inpainting, outpainting, 3D depth analysis, edge-based transformations, and artistic reimaginings. Photoshop apparently got bored and decided to become an art god and turned into Flux. With Flux, these guys are going to set the gold standard for AI image generation. If Adobe doesn’t step up, Photoshop might become the boomer tool of digital art, I tell ya!
Mistral’s launches new toys. Web search, canvas, and open-source goodies
Mistral tried copying stuff from OpenAI et al. and they released new features for their chat engine, Le Chat, including web search with citations, document analysis, and image generation. They’ve also open-sourced their latest models, because…. why nut?
What I think of this?
Mistral’s features are solid, but “Le Chat” is a crime against branding. If they want to compete with GPT-4 and Claude, they’ll need to offer more than cool tools with names that sound like a bad indie film.
SAMURAI: Zero-Shot Visual Tracking with Ninja-Level Precision
Researchers at the University of Washington developed SAMURAI, which is a motion-aware, memory-enhanced visual tracking system. It is based on Meta’s Segment Anything Model and performs zero-shot tracking without retraining. SAMURAI is basically a smart AI system for tracking objects in videos. If you’re watching a video where someone is running through a crowd, and you want to keep track of their movement…. enter the SAMURAI!
My take on this: This is the good stuff. Open-sourcing models like SAMURAI turns big AI breakthroughs into a communal upgrade. Let’s just hope it doesn’t end up being used to track you through your smart fridge.
For those who don’t want to read: Bugger off
Nah, just kidding. Here you go, mate:
Konnichiwa, watashi wa Maruko van Hyūne desu.
Elon Musk wants your X-rays for Grok
No kidding. Elon’s AI chatbot, Grok, now analyzes medical images because, sure, why nut?
Users are uploading X-rays en masse. Think of MRIs, and PET scans for early-stage diagnoses. Grok isn’t FDA-approved, but who needs that when you’ve got Musk’s seal of approval.
Hahahahahahaha.
Peeps are just gullible
🐑
My take: This idea wouldn’t last a day in the EU. Between GDPR, the AI Act, and the European Health Data Space, Grok would drown in red tape. But hey, Americans, upload your health data to a social media platform and see what happens. YOLO!
Videos of the Week
What’s got us buzzing? A mix of AI, tech, dystopia, and absurdity:
- The Rancher by Dave Clark
- A New Kind of Human by Alberto Papi Roldan
- Jag Max (Think Jaguar meets Mad Max). Really over the top AI-movie by Mark Day
That’s it for this week’s SundAI, without cream. Whether you’re grinding dungeons with Elon, letting AI write your next hit single, or just watching the chaos unfold, remember: the future’s coming fast, and it’s friggin’ weird as hell.
Signing off from the dystopian trenches of doom– Marco
Well, that’s a wrap for today. Tomorrow, I’ll have a fresh episode of TechTonic Shifts for you. If you enjoy my writing and want to support my work, feel free to buy me a coffee ♨️
Think a friend would enjoy this too? Share the newsletter and let them join the conversation. Google appreciates your likes by making my articles available to more readers.
Leave a Reply