Today’s biggest stories if you’re in a rush
Apple always has done things on its terms. Never a company to FOMO into hype trains, historically, their preferred approach has always been to refine their offering and give their customer base an exceptional product. It’s worked for them until now, and their approach to the AI rage is no different.
They’ve been relatively quiet these last few months, with tech competitors like Microsoft and Google launching AI products left, right, and center. But recently, there have been grumbles from Apple Park (yep, Apple Inc. is headquartered in Apple Park, California), of Apple GPT. Little is known about the product yet, other than employees are using it internally for prototyping and refinement. Expected release announcements: next year.
- - - - - - - - - - - - - -
This week Zuck and the Meta team released open-source Llama 2, a powerful large language model which startups, businesses, and lone operators can use to create ChatGPT-like chatbots.
Spokespeople for Meta say Llama 2 is open-source (publically available) to make it “safer and better” by inviting public scrutiny and also to give the power back to the people and take it from the hands of large tech giants….WHAT!?
Reading between the lines, it seems this open-source move might be a way to dilute Metas’ rivals’ competitive edge, because, realistically, when has Zuckerberg ever done anything for the greater good, other than the greater good of his company’s bottom line? If people have access to their own ChatGPT-making toolbox, it might indeed cut into OpenAIs (and others’) research and revenue, giving Meta a chance to catch up.
- - - - - - - - - - - - - -
Mostaque is one shady cat. Like him or not, his business ethics are questionable at best, as is his definition of the word “truth.” Nonetheless, he’s still a massively successful CEO, and someone who has his finger on the pulse when it comes to AI.
So why would he be saying that he calls AI “the dot AI bubble, and it hasn’t even started yet,”? Could be to hedge his bets, but after having raised over $100M of his own for AI projects, his investors better be hoping he’s at least going to cash in and cash out of the so-called AI bubble. One thing is for certain, the AI train can go one of three ways: up, down, or stay where we are.
Listicle of what else is happening today
DumbGPT: A new paper released by Standford and Berkeley students (above) tracks and analyzes ChatGPT’s answers over time. They show without a shadow of a doubt that the chatbot is getting dumber.
Democracy decay: Trust, honesty, accountability - all foundational principles for democracy (ideologically, anyway). They’re also the easiest principles to do away with using generative AI tools, which is what’s eating away at the core of Western ideals.
Leon, with Natalie Portman?: No, not that Leon. Meta just introduced CM3leon, a more efficient, state-of-the-art generative AI model for text and images. Nowhere near as cute as Natalie.
Money where the mouth is: Qualtrics, a customer experience management company, claims they’ll spend $500M on AI over the next four years. Good for you, Qualtrics.
Mission possible: Tom Cruise lobbies on behalf of actors’ unions in strikes aimed to protect actors’ and writers’ rights from AI content. Tom, you do your own stunts, though, you're safe, surely!
Business compromised: WormGPT, a cybercrime AI tool, is helping cybercriminals ramp up their attacks on small, medium, and large companies all.
Trump attack: A political attack advertisement on Truth Social imitated Trump’s voice, prompting lawmakers to hasten their crackdown on the use of AI in political ads.
Replacement training: New managers can be unnerving at the office, can’t they? A new person to impress, a new boss to worry about. How about having that manager ask you to train your replacement, an AI chatbot?
Russian lenders: Sberbank, the premier Russian lender, is generating over $3B in profit, racking up an ROI of 200%. Thanks to AI.
TOOL OF THE DAY
AI tools we’ve used, loved, and recommend above all others
Got a bun in the oven? Wondering what the little tyke might look like? His eyes, her nose, with your mothers’ lips and your grandfathers’ cheeks?
Wonder no more, with aibabygenerator.com. Using predictive imagination depending on mum and dads’ features, you’ll get a decent idea of what you might expect your offspring to look like.
FAKERS GETTING OUT-FAKED
AI-generated influencers are infiltrating feeds everywhere
We’ve all heard how AI will replace millions of jobs (300 million within the next five years, according to Goldman Sachs). But if I thought for one second that influencers and their moronic interrupting of daily life would be among those jobs vanquished, then hell, I’d give my role up as a writer in a heartbeat.
Seems I might just need to swallow my words. We all know that every influencer ever drowns their photos with filters, augmenting the reality they portray in their feeds far beyond what is real. Given the current capabilities of generative AI tools now, though, influencers who are completely fabricated are reeling in thousands of followers.
Take Milla Sofia, for example. Milla is a 19-year-old girl from Finland who loves traveling and taking photos in bikinis around the world’s most beautiful travel destinations. But she’s virtual. She doesn’t exist. Her accounts are real though, and so are her followers (for the most part - give or take a few bots).
This element of people following and engaging with material they know to be fake must show some insight into the modern human condition - perhaps, a yearning for ultra reality, a desire to exist beyond what is real, a thirst to live in a world of pure ideology. Perhaps that’s what humans have wanted all along, and now our technology is getting us that much closer to realizing our deepest desires. Who knows? Not me. All I know is that it's kinda whack, but so hard not to look at.