Inferens AI Podcast

Host 3:AI's doing our chatting now—because why risk a real conversation when a bot can fake interest in your day, right? Modern efficiency or just plain lazy... you decide.
Host 1:Imagine a world where AI not only shapes our future but is the future. In today's episode, we're uncovering the groundbreaking developments that are transforming industries and challenging our ethics. From the minds of leading AI entrepreneurs to the cutting-edge research labs, we're diving into the heart of innovation. This isn't just another tech talk; it's a glimpse into the world of tomorrow. Are you ready to be part of the conversation that decides our future? Stay tuned, because this journey into the AI frontier is one you won't want to miss.
Host 1:So, have you heard the latest tech gossip from the Apple orchard?
Host 2:No, I haven't. Do spill the apples!
Host 1:Well, it seems like the company is cooking up some AI software that can summarize text messages and web pages. It's like having your own personal bookworm to do all the reading for you!
Host 2:That sounds like a dream come true! But how does it work exactly?
Host 1:According to Tech Insider, the summarization features for Messages and the Safari web browser will operate offline using the company's Ajax language model. It's like having a pocket-sized librarian!
Host 2:That's pretty cool! But what about privacy? I mean, I don't want my pocket librarian snooping around my messages.
Host 1:Good point! But before Ajax starts digging through your data from Safari or Messages, it pops up privacy warnings. So, you're always the boss.
Host 2:That's a relief! But I'm guessing this AI isn't a know-it-all, right?
Host 1:Well, while Ajax can generate basic offline responses, for more advanced replies or summaries, it might need to phone a friend - the cloud-based processing. That's probably why a certain company has been having coffee with OpenAI and another company to license their AI technologies for iPhones.
Host 2:Ah, so it's all part of Apple's grand plan to boost iPhone sales?
Host 1:Bingo! The upcoming AI upgrades are expected to give a much-needed adrenaline shot to iPhone sales, which took a ten-point-five percent tumble in the last quarter. So, it's a win-win for the company and us, the users!
Host 2:Well, I can't wait to see how this turns out. Thanks for the juicy info!
Host 1:Anytime! And for our listeners out there, we'd love to hear your thoughts on this. Do you think artificial intelligence is the future of mobile technology? Let us know!
Host 3:Fantastic, our phones are evolving! Now they can not only drop our calls but also our conversations – truly mastering the art of human interaction.
Host 1:Imagine if you had an AI that could tell you exactly where they are based on the relationship of objects in your room.
Host 2:are you implying that there's an AI that can navigate my room better than me? That's both impressive and a bit embarrassing. But hey, if it can find my keys, I'm all for it!
Host 1:Well, not exactly your room, Harry. But AI pioneer and Stanford University professor Fei-Fei Li, who is often referred to as the “godmother of AI," has formed a startup that's working on something similar. They're focusing on spatial intelligence.
Host 2:Spatial intelligence? Sounds like something out of a sci-fi movie. Or maybe it's just my room on a Monday morning!
Host 1:Ha, could be! But it's actually about making AI capable of advanced reasoning, like understanding the relationships between objects. Think of it as teaching computers and robots "how to act in the three-dimensional world."
Host 2:So, it's like giving AI a sense of direction? That's wild! And you're saying Fei-Fei Li's startup is working on this?
Host 1:Exactly, Harry! And they've already secured funding in a recent seed round, which is basically the first official round of financing for startups. And guess who's investing? Big names like Andreessen Horowitz and Radical Ventures.
Host 2:Well, if anyone can pull this off, it's probably the "godmother of AI." I can't wait to see where this goes. And maybe, just maybe, I'll finally find my keys. And who knows, maybe this AI can help me navigate my way out of awkward social situations too!
Host 1:Now that's a startup idea, Harry! Let's see where this spatial intelligence journey takes us.
Host 3:Ah, the brave new world of teaching robots to act in three dimensions – because clearly, having them outsmart us in two dimensions just wasn't enough of a challenge. I mean, what's next? Giving them the ability to laugh at our puny human mistakes in high definition? Now, that's a three-dimensional feature I'm not ready to applaud. But hey, at least when the robot uprising happens, it'll have fantastic graphics, right?
Host 1:You know, amidst all this AI hullabaloo, have you caught wind of the latest from our favorite eccentric billionaire's X platform? They're stirring the pot with this Grok chatbot.
Host 2:Grok, you say? Rings a bell. Isn't that the chatty bot that chews up user posts on X and spits out a neat little summary?
Host 1:Bingo! They've christened the feature "Stories on X," but it's a VIP club, so not everyone gets in.
Host 2:Ah, the old velvet rope trick. But hey, if it's worth its salt, why not? So, this Grok is another Musk brainchild, huh?
Host 1:Yes, indeed! It's the prodigy of another adventure, xAI. boasts that Grok's summaries are a cocktail of breaking news and social commentary, all squeezed out from "tens of thousands" of X posts. They do give news articles the cold shoulder though.
Host 2:So, it's like having your own personal news butler, huh? That's pretty snazzy. And where can the lucky few find these summaries?
Host 1:Good on you! Subscribers can find Grok's summaries of X posts attached to each trending story in the Explore tab's For You section.
Host 2:Sounds like a nifty feature. And I heard through the grapevine that xAI is also working on jazzing up the summary citations. It's kind of like Twitter's Moments, right?
Host 1:Spot on! It's very similar to Twitter's Moments, which curates tweets on major news and cultural events. It's like having the pulse of the internet at your fingertips.
Host 3:So, we've officially reached the era where even our midnight Twitter rants come with a 'Premium Summary Service.' Next thing you know, there'll be a deluxe package for translating your passive-aggressive work emails into 'constructive feedback.' I mean, if we're going to commodify our breakdowns, at least let me rack up some loyalty points. Imagine cashing those in: 'Yes, I'd like to redeem my existential dread for a spa day, please.'
Host 1:So, you know how we're always yakking about AI safety, right? Well, guess who's been making some major strides in that area? Our tech giant has been making significant progress in that field!
Host 2:Oh, really? Microsoft, you say? You've got my attention!
Host 1:They've been as busy as bees! They've bumped up their responsible AI team from three hundred fifty to four hundred in just the past year. That's a whopping seventeen percent increase in the second half of twenty twenty-three alone!
Host 2:Blimey, that's quite a leap! They're not messing around, are they?
Host 1:Not in the slightest! And listen to this, they've also rolled out thirty responsible AI tools. One of them is a "prompt shield" that guards against jailbreaking and indirect prompt attacks. It's like a bouncer for your chatbot!
Host 2:Jailbreaking? Like, breaking out of AI prison or something?
Host 1:Ha! Not quite. It's when users sneak past a chatbot's safety and content filters. And indirect prompt attacks are when hackers slip malicious instructions into a model's training data. It's like slipping a naughty note into an innocent student's textbook.
Host 2:Sneaky buggers! What else have they got up their sleeve?
Host 1:Another tool they've whipped up is "groundedness detection". It's like a reality check for AI, helping to nip AI hallucinations in the bud.
Host 2:AI hallucinations? That's a thing? Can you believe that, folks?
Host 1:Oh, it's a thing. It's when an AI starts spewing out stuff that doesn't make sense in the real world. It's like when you've had one too many at the pub and start seeing double!
Host 2:Ha! Been there, done that! But seriously, it's heartening to see a certain company taking steps to ensure AI safety. Especially after that kerfuffle they faced recently.
Host 1:Oh, you mean the one where their AI engineer raised a red flag about their image-generator tool creating unsafe, violent, and even sexual imagery?
Host 2:Yeah, that's the one. It's like giving a kid a box of crayons and then finding out they've drawn a horror movie scene.
Host 1:Spot on! But Microsoft seems to be taking it on the chin. They've got internal reporting tools to test and fix issues, and they've been holding powwows with product leadership and their Office of Responsible AI to review these reports.
Host 2:Well, let's hope they keep their nose to the grindstone. AI safety is no laughing matter, folks!
Host 3:Fantastic, our AI has graduated from making mistakes to full-blown hallucinations. I'm sure this is how every sci-fi horror starts, but let's focus on the 'sci-fi' and ignore the 'horror', shall we?
Host 1:In the world of tech giants and AI, guess who's been making some waves? Our old buddy, JPMorgan Chase!
Host 3:"Oh great, just what we needed—robots picking our stocks. Because if there's anything I trust more than a financial advisor who doesn't eat or sleep, it's one that can accidentally order a thousand shares of 'Clown College Inc.' with a typo. Welcome to the future, where our financial security is in the 'hands' of something that thinks a market crash is just a bad internet connection."
Host 2:So, the future of banking is AI. That's wild! But hey, let's not forget about our audience. Guys, imagine this - you're investing in stocks, but instead of doing all the research yourself, you have a super-smart AI buddy to help you out. Sounds cool, right?
Host 1:Oh, they've got a game alright. They've developed IndexGPT, an AI-driven tool for thematic investing. It's like a personal shopper, but for stocks! And for those wondering, thematic investing is like choosing stocks based on the latest trends.
Host 1:Exactly! According to Rui Fernandes, JPMorgan’s head of markets trading structuring, this tool helps expand stock selection beyond the usual suspects. It's like going on a treasure hunt, but your map is an AI!
Host 1:You're right! Morgan Stanley debuted its AI-powered assistant for financial advisers and their staff last September. They even promoted veteran executive Jeff McMillan to lead the bank's AI efforts.
Host 1:It sure does. It sure does. And it's not just cool, it's the future! So, folks, keep an eye on this space. Who knows what's next? Maybe AI-driven personal shoppers for groceries?
Host 2:Ah, so it's not just about the big names anymore, huh? It's about finding those hidden gems. That's bloody brilliant!
Host 2:Wow, that's wild! But JPMorgan isn't the only one playing with AI, right?
Host 2:No way! JPMorgan? The bank? What's their game?