Inferens AI Podcast

Host 3:"Why bother with a dentist when you can just Google 'how to repair your own teeth'? I tried eating rocks to test it out—turns out, the internet's not always right. Who knew?"
Host 1:Ever pondered how AI's unexpected errors could redefine innovation or how its collaborations are shaping our future? In today's episode, we delve into the heart of AI's most intriguing developments and mishaps. From the labs of researchers to the strategies of entrepreneurs, we uncover the pivotal moments that are setting the course for tomorrow's AI. Get ready for an insightful journey through the complexities and triumphs of artificial intelligence. It's not just a ride; it's the frontier of discovery!
Host 1:So, have you heard about Google's new AI Overviews in its search engine? It's like a quick summary of the top search results.
Host 2:Text Host 2: Oh, you mean the one that's been churning out some pretty wild answers? Yeah, I heard about it. They rolled it out to U.S.-based searchers just last week.
Host 1:Yes, that's the one! It's been giving some pretty bizarre advice. Like, it suggested using glue to keep cheese on pizza and even recommended users eat at least one rock per day!
Host 2:(laughs) That's hilarious! But also kind of dangerous, don't you think? I mean, it gave incorrect advice on rattlesnake bites and made errors about the number of U.S. presidents. And it even falsely claimed former President Obama is Muslim.
Host 1:Yeah, it's definitely a concern. Some of the answers it's been pulling are from Reddit posts and even The Onion, which is a satirical news site.
Host 2:Text Host 2: Well, that explains the rock-eating advice! But seriously, it's a bit worrying. I mean, people rely on Google for accurate information.
Host 1:True, but the company has said that these mistakes are "extremely rare queries and aren't representative of most people's experiences." They've removed the overviews for some queries that produced incorrect information.
Host 2:Well, that's a relief. I mean, I don't want to be eating rocks anytime soon.
Host 1:(laughs) Me neither. Google also said they extensively tested the feature before launch and are "taking swift action" based on feedback.
Host 2:That's good to hear. And if anyone out there is still getting weird AI-generated answers, The Verge has shared ways to remove them from Google search results.
Host 1:That's a handy tip. So folks, remember to double-check any advice you get from Google, especially if it involves eating rocks or using glue on your pizza!
Host 3:Considering the latest diet trends, I've decided to take my nutrition advice from a geologist. Eating rocks daily? Finally, some solid advice in these turbulent times. Crunchy, full of minerals, and hey, it's gluten-free!
Host 1:my man, you're always up for some juicy tech gossip, aren't you?
Host 2:you know me too well! I'm all ears.
Host 1:Alright, brace yourself. You know how Meta and Elon Musk’s xAI company are like the Tom and Jerry of the tech world, right?
Host 2:Yeah, always at each other's throats, trying to outdo each other. What's the latest?
Host 1:Well, they're both vying for a partnership with Character.ai. It's like a high-stakes game of tug-of-war!
Host 2:Character.ai? Isn't that the startup that makes virtual versions of celebrities and historical figures?
Host 1:Bingo! They're the ones. They were founded just last year by a couple of ex-Google researchers. They're also the brains behind Google's LaMDA AI conversational tech. Pretty cool, huh?
Host 2:Damn, that's impressive. So, what's the deal with Meta and explainable AI?
Host 1:Well, according to some insiders, Meta has been in talks with Character.ai about collaborating on pre-training and developing models. And guess what? xAI has also been sniffing around for a similar partnership.
Host 2:So, it's a classic case of "anything you can do, I can do better"?
Host 1:Exactly! But as of now, no deals have been struck. It's still anyone's game.
Host 2:Interesting. And what about their own chatbots?
Host 1:Well, Meta launched its own celebrity-type chatbot personas, twenty-eight AI characters, across its messaging apps last fall. And xAI's chatbot Grok, launched in November, is available to X premium subscribers on the social media platform. They're even planning to make Grok multimodal, meaning it can generate text-based responses to images.
Host 2:Wow, that's some next-level stuff. It's like a sci-fi movie come to life!
Host 1:You said it! It's a thrilling time to be alive in the tech world. Let's see who wins this round of the AI tug-of-war.
Host 3:Ah, just what the world was waiting for—another batch of virtual Kardashians. Because, clearly, what we lack in reality is more reality TV stars in our digital lives.
Host 1:Spot on! It's not available for public use yet, but it's operational in a research center. It can answer questions, generate reports, summarize info, and even translate between Chinese and English. It's like a super intelligent parrot, but instead of "Polly wants a cracker," it's "Polly advocates socialism with Chinese characteristics."
Host 1:Well, according to PitchBook's latest data, AI and ML venture capital funding in the first quarter of twenty twenty-four reached a whopping twenty-one point six billion dollars. That's a slight dip from the previous quarter's twenty-two point three billion dollars, but still a hefty sum, don't you think?
Host 1:You're spot on! You've been doing your homework, haven't you? And it's not just them. DeepL just raised three hundred million dollars, Scale AI secured one billion dollars, and a French startup called H raised two hundred twenty million dollars in a seed round. It's raining money in the AI world!
Host 1:my man, you're always up for some AI chit-chat, aren't you? Well, buckle up, 'cause this one's a doozy! China's internet watchdog, the Cyberspace Administration of China, has developed an AI chatbot trained on President Xi Jinping's ideology. It's like having a mini-Xi in your pocket!
Host 1:Oh, you're such a riot! Yes, they've got an AI assistant. They first let the cat out of the bag in September. It's built on Host's Llama three model. And in April, they unleashed it to millions of users across Facebook, Messenger, Instagram, and WhatsApp.
Host 1:You're right! It's like we're living in an episode of Black Mirror. But hey, that's the power of AI for you! And to our listeners out there, what do you think about this? Would you want a mini-president in your pocket? Let us know!
Host 1:Absolutely! Xi Jinping Thought is a set of policies and ideals that guide the Chinese Communist Party. It's all about socialism, the absolute authority of the Party, and a few other principles. Fourteen in total, to be exact.
Host 1:It sure is. But remember, with great power comes great responsibility. These companies have a lot of potential, but they also have a lot to prove. So, let's see how they use this funding to push the boundaries of what AI can do.
Host 1:Hey, you know how we're always trying to outsmart Siri and Alexa, right? Well, the company is stepping into the ring. They're cooking up a paid version of their AI assistant. Got this hot gossip from an internal post.
Host 3:"Ah, the era where AI not only debates politics but also leads the charge at robot rallies. I can see it now: a world where our biggest political debates aren't between left and right, but between on and off."
Host 1:A significant chunk of last quarter's venture capital funding, about five point three billion dollars, went to what they call "megadeals" for foundation model companies like Anthropic, Mistral AI, and xAI.
Host 2:Absolutely. And folks, remember, we're just the messengers here, sharing the news. We're as excited as you are to see where all this leads. So, stay tuned and keep buzzing with us in this AI hive!
Host 1:You're on the money. But Meta's keeping mum on the price. It's all hush-hush for now. But you know, as soon as we sniff out the details, we'll be spilling the beans to all our listeners out there!
Host 2:Wait, isn't xAI that generative AI startup? I heard they're about to close a six billion dollar funding round soon. That's gonna skyrocket their valuation to eighteen billion dollars, right?
Host 2:No kidding! I've been hearing about these massive funding rounds. It's like every tech nerd's dream come true. But do spill the beans, how much are we talking about?
Host 3:"Oh fantastic, another AI funded by a billionaire's pocket change. Because clearly, the solution to all our problems starts with 'Siri, take over the world.'"
Host 3:Fantastic, we've now monetized our digital moans with paywalled chatbots. What's next in this dystopian comedy? Shelling out coins to prove we're not robots?
Host 1:Well, they're planning to charge for the fancy features of the chatbot. The nitty-gritty is still top secret, but it's got the tech world in a tizzy.
Host 2:Wow, that's some next-generation stuff! And it's trained on Xi Jinping Thought, right? That's the president's doctrine that's taught in schools?
Host 2:What the...! Are you pulling my leg? The company has an AI assistant? I thought they were all about poking friends and virtual hangouts.
Host 2:I'm itching to know the price though. I mean, three tech giants each charge twenty smackers monthly for their top-notch chatbots.
Host 1:You know what's been buzzing lately? The AI and ML scene. It's like a beehive of activity, with money flowing in like honey!
Host 2:And the bot's being used for cybersecurity and IT research, right? That's like something straight out of a sci-fi movie!
Host 2:Damn! It's like every other day there's a new AI startup popping up with a billion-dollar funding round. It's crazy!
Host 2:Holy guacamole! That's massive. But hold on, they're going to put a price tag on it? What's the scoop on that?
Host 2:No kidding! So, this bot, it's like a digital version of the president, echoing his ideology?
Host 2:Holy guacamole! That's a lot of money. But where's all this moolah going?