Episode Categories:
Resources:
Join the discussion: Ask Jordan questions on AI
Upcoming Episodes: Check out the upcoming Everyday AI Livestream lineup
Connect with Jordan Wilson: LinkedIn Profile
Try Our Free AI Prompting Course: Register for our free Prime, Prompt, and Polish AI Course!
AI News That Matters for Business – September 23, 2024
The world of artificial intelligence is moving at breakneck speed, and it's crucial for businesses and investors to stay up-to-date. In this article, we will discuss the most important developments from September 23, 2024.
AI Applications in Hollywood
Artificial intelligence (AI) has become a game-changer in the film industry. Several Hollywood studios are deploying AI models to create storyboards, animations, and special effects. While legal limitations prevent replicating an actor's likeness without consent, AI technology continues to pave the way for efficient and innovative filmmaking.
Runway and Luma Labs
Runway and Luma Labs are accelerating the evolution of AI in video generation. Their newly released APIs, featuring the Runway Gen 3 Alpha Turbo and Luma’s Dream Machine models, are primed to activate a slew of AI video tools, transforming the future of digital content creation.
Partnership and Investment
In a landmark deal, Microsoft and BlackRock have announced a $30 billion investment partnership to enhance US AI competitiveness. The partnership will focus on developing AI data centers and energy projects, with an overarching goal to mobilize $100 billion in total investments, consolidating America’s position as a global AI powerhouse.
Energy Consumption Concerns with AI
However, the growing demands of AI come with significant power consumption concerns. Some firms are reporting up to a 50% surge in power usage due to increased AI activity. Sustainable energy solutions will become increasingly crucial in this AI-expanded world.
Alibaba’s Quinn 2.5
Chinese tech giant Alibaba continues to make strides in AI, launching over 100 AI models. Among these is the impressive Quinn 2.5, which outperforms Meta’s Llama 3.1. This range of models presents diverse applications, especially excelling in coding and mathematics.
Salesforce AI Training Initiative
In response to the growing AI skill gap, Salesforce has launched the AI for All initiative, investing $50 million in AI education. The initiative offers free AI courses and will establish a physical AI training hub in San Francisco. These efforts are key to fostering new talent and preparing future workforces for an AI-dominant landscape.
Global AI Safety Summit
Looking towards international collaboration, the US is set to host a global AI Safety Summit in November. A host of AI experts and government scientists will convene to foster safety, security, and trust in the field of AI.
AI Legislation on Deepfakes
Legislation has caught up with technology with California implementing several laws addressing deepfakes, particularly in political advertising. These laws aim to maintain electoral integrity, mandating disclosures if adverts contain AI-altered material.
OpenAI Funding
OpenAI is planning a game-changing $6.5 billion fundraiser, led by Thrive Capital, with potential backers including tech giants like Apple, NVIDIA, and Microsoft. With such substantial backing, OpenAI is set to reframe the AI landscape.
Apple's AI Advancements
Keeping up with this rapid AI innovation, Apple has released a public beta for its latest operating systems featuring advanced AI capabilities. These new tools promise to redefine the way users interact with their devices, bolstering AI’s integration into quotidian life.
AI’s Impact on Jobs
With AI's growing involvement in our everyday lives, it's predicted that this technology may replace more jobs than it creates. It underscores the importance of embracing AI and exploiting available training opportunities to remain competitive in the job market.
OpenAI's Innovative "o one" Model
Finally, offering additional proof that AI is the future, OpenAI unveiled its "o one" reasoning model. Not part of the GPT series, this innovative model potentially demonstrates OpenAI's second stage of AI development: reasoning capability.
The integration of advanced AI technology is undeniably shaping our world, from the film industry and law to education and job markets. It is imperative that businesses adapt and integrate these rapid advancements, allowing AI to enrich their operations and procedures. Stay tuned for more AI news and insights that can help shape your business trajectory.
Topics Covered in This Episode
1. AI Partnerships and Investments
2. Salesforce AI Training Initiative
3. White House AI Global Safety Summit
4. California's AI Legislations
5. Lionsgate and Runway Partnership
6. OpenAI new Model and Funding
7. Apple Intelligence
Podcast Transcript
Jordan Wilson [00:00:18]:
What is OpenAI raising $6,500,000,000 for? Apple Intelligence is kind of out, and California and the White House are trying to lead on AI safety, but will it even matter? We'll tackle those topics today and a whole lot more on everyday AI. What's going on y'all? My name is Jordan Wilson, and I'm the host of everyday AI. Before we get started, though, a quick word from our partners at Microsoft. So why should you listen to the WorkLab Microsoft the WorkLab podcast from Microsoft? Because it's made for leaders who know they must adapt to stay ahead. WorkLab is the place to find real world lessons and actionable insights to prepare you for the next phase of AI at work. That's Worklab, no spaces, available wherever you get your podcasts. Alright. Speaking of podcasts, this is for you.
Jordan Wilson [00:01:17]:
This is Everyday AI is your daily livestream podcast and free daily newsletter helping us all learn and leverage generative AI. So if you are listening on the podcast, trying to grow your company and career, thank you for joining. Make sure to check out your show notes for more as well as a link to our website at your everydayai.com. There you'll find a spot to sign up for our free daily newsletter as well as literally hundreds of episodes. It is a free generative AI university that you can go read and learn at any time. So thank you for joining into our livestream audience. Thank you all for tuning in as well. Rolando, Joe, and Jay, Brian, Tara, Michael, everyone else, thank you for tuning in.
Jordan Wilson [00:01:58]:
You know, you can spend hours literally every single day trying to keep up with AI news. I mean, we do it, so you don't have to. And on Mondays, most Mondays, we do our AI news that matters. This is your one stop weekly recap to not just get caught up on what's going on in the world of AI news, but how it applies, to your company or career and how you can get ahead. Alright. So let's get straight into it and talk about that huge number, $6,500,000,000. So according to reports, OpenAI is expected to raise approximately $6,500,000,000 at a staggering pre money valuation of a $150,000,000,000, marking what could be the largest venture capital round in history. Alright.
Jordan Wilson [00:02:46]:
So, this funding is notable as it surpasses the previous record reportedly of $6,000,000,000 raised by Elon Musk's x AI earlier this year. Yeah. And we've, seen obviously Elon Musk, former cofounder of OpenAI, getting in a couple, spats both publicly and seemingly through lawsuits with OpenAI. So this, new funding round comes amid a backdrop of immense interest in generative AI with the company reportedly turning down 1,000,000,000 of dollars in oversubscribed or overcommitted funding. So Thrive Capital is leading the investment, contributing just over 1,250,000,000, which includes funds from a special purpose vehicle. Other potential backers include major players like this is oh, this is who's who of AI right here. Apple, NVIDIA, and obviously Microsoft indicating strong confidence in OpenAI's future prospects. The minimum investment required to participate in this round is reportedly set at $250,000,000 highlighting the high stakes involved.
Jordan Wilson [00:03:54]:
This funding round's valuation of a 150,000,000,000 is compared to the entire US VC market's total under management just about 25 years ago in 1999, which contributed to the Internet bubble. Yeah. So it's more than, all of that. That's wild. Even for more context, just look at 10 years ago. So the amount that OpenAI right now is aiming to raise in this round is equivalent to the total funding received by all startups in New York, Texas, and Florida combined 10 years ago. So OpenAI is raising money as previous reports have indicated the company is on pace to lose up to $5,000,000,000 this year. So, yeah, that was a single report, and we haven't really seen a lot outside of that.
Jordan Wilson [00:04:42]:
So is that the reason they're raising $6,500,000,000? Well, I mean, the reality is just about everyone is trying to throw money at OpenAI, and that is why the funding round is oversubscribed by 1,000,000,000 of dollars, and that's led OpenAI to implement a stringent selection process for investors. So the company is scheduled to notify yeah. It's like, you know, it's it's like winning a lottery apparently. Right? OpenAI is choose choosing, like, yeah. We'll take your 1,000,000,000. We won't take your 1,000,000,000. So, the company is scheduled to notify potential backers today about their, participation status. So, yeah, pretty, pretty wild.
Jordan Wilson [00:05:23]:
Right? And we're gonna talk about their new o one model a little bit later, but, you know, the timing of that. Right? Because we've heard about this, potential huge record breaking, funding round now for a couple of months. But obviously, the timing, they were kinda waiting until, seemingly until releasing their o one model, which we're gonna talk about here in a little bit. Alright. Speaking of big companies trying to make a splash with AI, Apple has kind of just released Apple Intelligence. Well, at least the public beta for its latest operating systems, marking a significant step in integrating advanced AI features into its devices. So this is pretty newsworthy as this now is officially opening up generative AI capabilities for potentially now billions of iPhone users. So let me just preface this.
Jordan Wilson [00:06:21]:
This is a beta. So they just released this, over the weekend. So, you do have to access and sign up for the beta program. So this isn't a normal you know, don't don't go, you know, checking your phone and say, oh, I have, AI now on my iPhone. It's not like that. So, it is just the public betas of iOS 18.1, iPadOS 18.1, and then macOS Sequoia 15.1. So a lot of new, at least new tools for Apple when it comes to AI, like text rewriting tools and a, revamped Siri design. Also, there is a new cleanup tool for photos that have been introduced allowing users to remove unwanted objects from their photos, enhancing the photo editing experience.
Jordan Wilson [00:07:12]:
So to access these bay, betas, users must register on Apple's beta software program site. I believe that's beta.apple.com, enabling them to download and install updates directly from settings. So, yes, you do have to, sign up for the beta program and then sign up for the early access to Apple Intelligence. So, yeah, if if if you, hear or see people, you know, there's probably gonna be a lot of news articles now because this is the first time, that users will actually have access to Apple Intelligence. But not everyone, just those who are signing up for the, beta program developers as well. So, here's the other thing y'all, and we've talked about this on the show a lot. There's literally nothing new here. Right? These are all, you know, I'd say probably the the the 2 or 3 biggest, you you know, updates coming with this Apple Intelligence are kind of, the ability to improve your written text kind of wherever you're at.
Jordan Wilson [00:08:10]:
I would say the, ability to summarize and prioritize, messages and notifications. And and then I mean, there's there's a lot of other, you you know, Siri improvements, but not everything is rolling out in this 18.1 update. So we will see. I didn't get a chance to do this myself, but we will see what's kind of available now, probably in the newsletter by the time you read it today versus what will be in the next beta. And I believe that one will be in, October. And then most updates, at least here in the US, will be rolled out, I believe, by February of 2025. So, yeah, very slow rollout here for Apple Intelligence. But, y'all, it's out.
Jordan Wilson [00:08:53]:
Go play with it. See if this changes, your workflow, but you do for the iPhone variation, you do have to have at least an iPhone 15 Pro or newer, or the new iPhone 16 that was just released a few days ago. So, if you are on the Mac looking at the Mac, operating system beta for Apple Intelligence, you do have to have a computer with an M1 Chip or higher, and then obviously update to your software. So yeah. I'm curious. Yeah. Douglas is saying here, you're a Mac guy, Jordan. Are you signed up for the beta? Yes.
Jordan Wilson [00:09:30]:
I am signed up. I think I was preparing for my show last night, and I I have a couple different computers that I use. So, yeah, it did finish downloading, and then I was so tired. So I'm like, yeah. I'll I'll take a look at this tomorrow after, doing the show. So, yeah, I'm I'm curious if if any of you out there have already, dived into this, Apple Intelligence beta. Alright. So the Biden administration is set to host a global AI summit on safety in November.
Jordan Wilson [00:10:01]:
So pretty big news here. The White House is set to host a pretty significant global safety summit on AI in November, reflecting the urgent need to address the evolution of AI technology and its obviously associated risks. So the summit will take place in San Francisco on November 20th 21st, co hosted by US Secretary of Commerce, Gina Raimondo, and Secretary of State, Anthony Blinken. So government scientists and AI experts from at least 9 countries and the EU are expected to participate, signaling a collaborative international effort to tackle AI challenges. So this meeting aims to establish safety, security, and trust in AI with Raimondo emphasizing the need for rules that govern AI usage to, be based on those principles. So representatives from the US, UK, Australia, Canada, France, Japan, Kenya, South Korea, Singapore, and the UE will reportedly, attend, but China, a major player in AI development, is notably absent from that list. The timing of this summit is obviously significant as well occurring just shortly after the US presidential election, which I believe is on November 5th, and ahead of a larger AI summit in Paris scheduled for February. So a key focus of the summit will be the establishment of a global AI incident response network designed to rapidly address and contain potential AI system failures or unintended consequences across international borders.
Jordan Wilson [00:11:40]:
Don't don't get this, don't mistake this y'all. This isn't laws. This isn't legislation. Right? All this is this is a public front by the US government to say, yeah. We're doing something about AI. That's the thing that I think a lot of people don't understand. I mean, whether you're from the US or not. I was a former journalist.
Jordan Wilson [00:12:01]:
I covered politics, you know, at a at a pretty, pretty high level. And people don't understand. For for any meaningful legislation to come out, of the US, it is probably going to take minimum 5 years. I would say probably a lot longer. So, you know, don't expect any new laws, rules, or regulations. This is just kind of, you know, world superpowers coming into coming together and, you know, kind of having unofficial rules. Right? So they're they're they're kind of writing an unofficial rule book that they hope everyone will follow, but these are not laws or actual regulations that are going to be put in place anywhere. The WorkLab podcast from Microsoft is made for leaders who want to understand how work is changing.
Jordan Wilson [00:12:51]:
Effective leaders adapt. They stay ahead of trends. They embrace any edge they can get. They also know that AI powered organizations will be better at spotting opportunities, creating new products and business models, and maximizing value. For real world lessons and actionable insights to help you stay ahead, check out the WorkLab podcast. The new season launches September 12th. That's w o r k l a b. No spaces.
Jordan Wilson [00:13:18]:
Available wherever you get your podcasts. Ah, sip of water. That's fresh. Alright. Speaking of California and government trying to take bold steps against AI, California governor Gavin Newsom has signed, well, I think 9, bills in total, but 3 pretty new ones aimed at addressing the growing threat of artificial intelligence generated deep fakes in political advertising, particularly as the 2024 election approaches here now in what? We're a a month and a half away from elections here in the US. So this move underscores the urgency for protecting electoral integrity in an increasingly, crazy world with AI and deepfakes. So a new law prohibits the creation and distribution of deepfakes related to elections within a 120 days before election days 60 days after, allowing courts to halt distribution and impose civil penalties. Well, why is this important? Well, all of the main social media networks are based in California.
Jordan Wilson [00:14:35]:
So large well, not all of them, but most of them and tech companies as well. So large social media platforms will be required to remove deceptive content under a under this groundbreaking law set to take effect, marking California as the 1st state to implement such regulations. Also, political campaigns must disclose if their advertisings contain AI altered material promoting transparency. So this legislation is obviously timely, and it follows Newsom's commitment to tackling election deepfakes after a lot of controversy. So a controversial video from Elon Musk, obviously the owner of Twitter now x, that featured altered images of vice president Kamala Harris, as well as former president Donald Trump, releasing, similar, manipulated with AI photos of his opponent in Kamala Harris. So the California law's unique provision extending protection for 60 days after an election addresses concerns also about post election misinformation campaigns, which have become increasingly prevalent in recent weeks. Gosh. It's gonna be a mess, so buckle up.
Jordan Wilson [00:15:47]:
Alright. So, Joni Ive and Sam Altman, new details about their collaboration on an AI hardware project and potentially raising $1,000,000,000. So, Joni Ives partnership with OpenAI CEO, Sam Altman, and and Joni is the, you know, famed Apple iPhone designer, if you don't know. Also, I I feel I'm always mispronouncing, his name. Anyways, so Ive's partnership, reported partnership with OpenAI Sam, CEO Sam Altman is marking a pretty significant development in the intersection of design in AI, drawing attention from the tech community on what this new AI enabled hardware project could look like. So Ives has confirmed he is collaborating. So, yeah, we've just heard a lot of reporting dating back to last year. But finally, he has confirmed in a New York Times, piece that he is collaborating with Sam Ullman on a new AI hardware project.
Jordan Wilson [00:16:48]:
So reportedly, this project, they've already received a $1,000,000,000 in funding and the team is up to 10 employees, including key figures from the original iPhone development teams. So Love From is Ives' design firm, and it is leading the project which is based in a 32,000 square foot office in San Francisco. So not exactly y'all 32,000 for 10 employees. So that makes me wonder if they're actually going to be, doing anything with hardware in there. That's a huge building for 10 people. And also discussions between Ive and Altman have centered on how generative AI can create a new computing device that surpasses traditional software capabilities. Previous rumors suggested the device could draw inspiration from touchscreen technology and the original iPhone, although specific details remain vague. So, yeah, I don't know what this device could be.
Jordan Wilson [00:17:47]:
Right? We've seen the AI hardware devices so far do really bad. Right? So, the humane pin has reportedly flopped. The rabbit r one not didn't get good reviews. So I'd say aside from Meta's collaboration with Ray Bans on the, Meta AI glasses, I don't think a lot of AI enabled hardware has fared significantly well. So, should be interesting, what form factor this device might be. Hey, live stream audience or or podcast audience. Right? If you're if you're listening on the podcast, I always put my email in my LinkedIn, in there. Let me know what do you think.
Jordan Wilson [00:18:31]:
What is this new, AI powered hardware going to be? It's not gonna be a phone. Is it just gonna be a device? I don't know. To me, even though I love AI, I love technology, I don't see myself carrying a a third party device unless it's something that just is easily, you know, attached to my home computer. Otherwise, I don't really see a need for it or really a use case if I'm being honest. Alright. But something that I do have a use case for, I love the new episode. If you if you all didn't listen to it, you gotta go check it out. But a quick word from our partners at Microsoft and the WorkLab Podcast.
Jordan Wilson [00:19:10]:
So the WorkLab Podcast from Microsoft is made for leaders who want to understand how work is changing. Effective leaders adapt, they stay ahead of trends, and they embrace any edge they can get. They also know that AI powered organizations will be better at spotting opportunities, creating new products and business models, and maximizing value. So for real world lessons and actionable insights to help you stay ahead, check out the WorkLab podcast. That's worklab, no spaces, available wherever you get your podcasts. Yeah. They just launched, the new season. It was fantastic.
Jordan Wilson [00:19:50]:
So make sure you go give that a listen. Alright. Listen to this huge partnership between Lions Gate and Runway. So in a significant move for the film industry, major Hollywood studio Lionsgate has formed a partnership with AI firm Runway to create a custom model trained on Lionsgate's extensive film and TV archives. So this collaboration represents a growing trend among Hollywood Studios to leverage AI technology for production efficiency. So this partnership is the first of its kind, at least publicly announced focused on a developing focused on developing a video model exclusively for Lionsgate's use. So the deal involves fine tuning video generation models with Lionsgate's owned content primarily for pre visualization and also complex visual effects. Unlike typical data licensing agreements, this partnership allows Lionsgate to maintain control over its IP while enhancing production capabilities.
Jordan Wilson [00:21:02]:
So, current applications of these AI models are expected to include generating storyboards, animations, and complex special effects, which can be costly and time consuming. The deal, however, does not include replicating actors' likenesses, adhering to legal considerations regarding actor consent and IP rights for, those actors, which we just talked about that on the show last week, some new, laws from California. Also, reportedly, 6 other major Hollywood studios are reportedly pursuing similar fine tuning initiatives, indicating a broader industry shift toward AI integration. Yeah. Y'all, I told everyone about this a year ago, but well, not about this one, but I said AI generated video, literally in 2023. I talked about my my bold predictions for 2024, and I said AI powered video is going to be mainstream. It's it's going to be, coming to theaters and and Netflix and everywhere that you watch your videos in 2024. And at the time, the technology was not there, and I don't think most people believe me, but here you go.
Jordan Wilson [00:22:13]:
A name brand studio already entered into a partnership, and presumably, they're going to start, using this technology on films this year. So, pretty big news. And speaking of news with Runway, so both Runway and its closest competitor, I'd say, Luma Labs, have announced their APIs for videos. Also, Runway announced their video to video model, and it is so good. Alright. But both Runway and Luma Labs are stepping up the game in AI video generation by releasing APIs for their latest model. So that's Runway's Gen 3 Alpha Turbo and Luma's Dream Machine. So this limited access for API could soon unlock a flood of new AI video, applications, enabling developers to create tools that generate bespoke video content with ease.
Jordan Wilson [00:23:11]:
So essentially, this is like rappers. Right? You've seen all these, you know, GPT rappers over the last couple of years. Right? Where essentially other companies slap a new UI, a new UX, on a certain technology and, you know, they hope to call it a new product. So that's what we're seeing here with Runway and Luma Labs. So in the coming months, when you see probably, I would assume dozens of new AI tools saying, new video tool. It's probably just actually Runway or Luma Labs. So this is the first time one of the big players in the AI video, kind of scene has announced general API access. Speaking of partnerships, here's a $30,000,000,000 one.
Jordan Wilson [00:23:59]:
So, a significant partnership has been formed between Microsoft and BlackRock. So a $30,000,000,000 AI infrastructure investment partnership to create the Global AI Infrastructure Investment Partnership. I don't know if that's g a I I p or if it's we're gonna have to just spit out that whole phrase. So this new partnership aims to bolster American competitiveness through substantial investments in AI infrastructure. So this initiative underscores the growing confidence in AI technology and their potential on the economy. So the apart the partnership has established a $30,000,000,000 fund dedicated to developing essential infrastructure for AI, including energy, energy projects in data centers. So this is also, a little timely here. Right? Because we talked about on the show, last week as well.
Jordan Wilson [00:24:56]:
Microsoft reportedly revitalizing the 3 Mile Island Nuclear Plant. So, yeah, a big, big kind of partnership here to go all in on these data centers. So the data centers are crucial for the AI economy, and will be a primary focus of this new $30,000,000,000 partnership. And also the goal here is to eventually mobilize a $100,000,000,000 in total investments in this, in this project, which includes debt financing to support new AI infrastructure initiatives. So the rapid growth of computing requirements for AI has also caused some concerns about energy consumption and emissions with some firms experiencing increases of up to 50% of power consumption. So, yeah, we've been talking about this here on everyday AI for a very long time. But generative AI and and and the GPUs that all of these systems need. Right? So whether we're talking about ChatGPT or Runway or Gemini or maybe, you know, your company is using some sort of, proprietary, AI model or building your own.
Jordan Wilson [00:26:06]:
It all requires a lot of GPUs and powers and inference to train all these models. Right? So, power is going to be something that is going to become increasingly talked about, through the rest of 2024. Alright. Will we be talking about this new model, Quinn 2.5? So, Alibaba last week launched more than a 100 AI models. Yes. More than a 100. But the big one here is the Quinn 2.5 series that is a groundbreaking collection of 13 advanced AI models that really aim to redefine open source capabilities. So this new series has already outperformed Meta's Llama 3.1 in key benchmarks.
Jordan Wilson [00:26:56]:
So, Quinn from, Quinn from Alibaba is really trying to now be the leader in open source AI. So the Quinn 2.5 series features 13 separate models that cater to various applications, including coding and mathematics. That's really it looks like at least looking at the benchmarks that wears that's where this new model really thrives. So notably, these new models from the Chinese firm Alibaba have surpassed Meta's llama 31405 b. So essentially, Meta's top open source model. So it's surpassed that and has shown superior performance on the LiveBench AI benchmarks. So these new models from Alibaba are available in sizes ranging from 1,500,000,000 to an impressive 72,000,000,000 parameters, providing options for different user needs and computational resources. So, yeah.
Jordan Wilson [00:27:57]:
That's gotta be the, the highlight here. So we'll see how, Meta responds, and we'll see once other benchmarking services, go through and start benchmarking this new Quinn, 2.5 model, and to see if across the board, if it is going to be the most capable model, at least in terms of benchmarks for an open source model. So right now, that title has gone to, from Facebook parent company Meta, and their Meta llama 3 1405 b. But looking, the fact that reportedly a 72,000,000,000 parameter model And Quinn 2.5 is outbenching a 405,000,000,000 parameter model. It's pretty noteworthy. So, yeah, this is early so far, so we'll see if this holds true on all other benchmarks. But at least, with the service LiveBench AI, this new model pretty far ahead right now. Alright.
Jordan Wilson [00:28:54]:
Salesforce, y'all. So Salesforce is launching a free AI training initiative with a $50,000,000 investment, and this is open to all. So Salesforce has announced a significant move to enhance AI skills in the workforce by unveiling its AI for all initiative, which promises free AI training opportunities. So this initiative is pretty newsworthy as it just follows, you know, Salesforce just had their Dreamforce conference, and they announced their Agent Force. So, hey, there's another one I came up with a hot take on last year saying that there's gonna be more AI agents than humans this year. And, hey, the c the Salesforce CEO, himself said he foresees more than a 1000000000 agents just from Agent Force. So every time y'all think I'm crazy, you know, one of the biggest companies in the world, a year later, comes and says, oh, yeah. You you know, yeah, we're gonna have a 1000000000 agents.
Jordan Wilson [00:29:55]:
So, Salesforce will waive all fees for its AI for all, for its instructor, instructor led AI courses and credential programs on its trailhead training program until the end of 2025. So, yeah, you gotta you gotta go 15 months if you wanna go, dive into that AI for all training. But, hey, why wait if it's, free and available now? So Salesforce plans to also establish a physical AI training hub at its San Francisco headquarters dedicated to upskilling employees in AI tools and technology. So this training initiative represents a pretty substantial investment of $50,000,000 aimed at supporting communities and addressing the growing demand for AI expertise. Yeah. You know, upskilling, reskilling, cross skilling, skill sharing. Right? All these skill acronyms, they've really been the buzzword of 2024 because when we talk about, the potential disruption that AI has on jobs. Right? Even agent force.
Jordan Wilson [00:30:57]:
Right? We talked about this last Monday on the AI news that matters, but very seemingly impressive, technology that Salesforce is rolling out where essentially different agents, you can put them on your website, It can connect to your customer data. You can build guardrails. It is a low code, and no code solution, and it can handle inquiries seemingly very well. So you might think, like, okay. Well, what's gonna happen to potentially all these, you know, customer support, marketing, sales jobs? Well, we're not sure, but what we do know is upskilling and rescaling is going to continue to be one of the most talked about facets, I'd say, for the next 6 months because y'all follow the money. I'm not a pessimist. I am a realist. I I I follow the trends.
Jordan Wilson [00:31:39]:
I follow the money. I I personally think AI is going to, take away, quote, unquote, take away more jobs than it will ultimately create. So if you are listening now, don't worry. Just by tuning in every day, by reading our newsletter, you're ahead of your peers, but you also need to be training yourself. Take advantage of all these free training opportunities while they exist. We had the, founder of Grow with Google on the show last week talking about some of their free training resources as well. OpenAI OpenAI as well. Just, literally minutes ago, announced a new training collaboration as well with their OpenAI Academy.
Jordan Wilson [00:32:19]:
So go take advantage of all of these free resources right now. Alright. And let's wrap it up. Speaking of OpenAI, gotta save the best for last. Right? Gotta save the juicy stuff for the end. So OpenAI's Sam Altman has announced some major advancements at least with their own internal metrics on their new o one AI model. So, yeah, that's the model that was just released, about 2 weeks ago. So, you know, originally, this was called QStar, then it was called, strawberry, but now the official name we know is o one.
Jordan Wilson [00:32:58]:
So OpenAI did just release this more reasoning model, and y'all, please, everyone's getting this wrong. This is not GPT o one. This is just called open I oh, OpenAI o one. So right now, there is the o one preview and the o one mini, that is available for paid subscribers and depending, on what other tools you use. A lot of third party tools have started to roll out this, new model in their APIs as well. So here's what's different. Right? And I need to call this out because, you know, even people who are quote unquote AI experts are getting this wrong. This is not a GPT model.
Jordan Wilson [00:33:37]:
This is 2 different classes. So we will continue to see updates, to open AI's GPT 4 o model. So, right, whether we see a GPT 45, maybe GPT 5 in the coming, weeks or months, we'll see. There's been some rumors. We might see some updates on something tomorrow, with a voice, advanced, voice mode. So we'll see if that actually comes to fruition or not. But, these are 2 separate classes. You have your GPT class.
Jordan Wilson [00:34:07]:
And right now, at least what it's called is you have your o one class, this advanced reasoning model. Two different things. Two different classes. So, Altman was at a T Mobile event where OpenAI and T Mobile talked about some different partnerships. But I'd say the biggest takeaway from this is Sam Altman did essentially admit, that with the o one model, they've reached their own internal kind of stage 2. So OpenAI a couple months ago, whether you wanna say leaked or released, but there are 5 different stages or 5 steps to AGI or artificial general intelligence. Right? That is when the average, AI model is smarter and more capable than the smartest human beings. And essentially, you know, Sam Altman in this interview with T Mobile did say, yeah.
Jordan Wilson [00:34:56]:
This is stage 2. This is the reasoning level, you know, out of their 5 stages. So Altman outlined previously the 5 levels of AI development, essentially saying that the new o one model, this strawberry, is now at level 2, or reasoners. And then with the potential for the next level, level 3, which is agents to emerge shortly. Right? He essentially said once you kinda once we, reach this level 2, which, you know, level 2 in theory took a while, took a couple of years. He did say that reaching level 3, which is capable agents, will emerge shortly. So here's here's the thing y'all. This new model, o one, it is completely different.
Jordan Wilson [00:35:43]:
Right? There is no comparison. Yes. I think we compare it to other models out there, you you know, maybe comparing it to, models from Anthropic, such as Claude 35 SONNET, comparing it to models such as Google Gemini, their their one point five. These are different things. Right? I I I almost wish that it wasn't in even in the same benchmarking programs because it is a new class of model. It is a reasoning model. So even, right, we teach, prime prompt polish. Right? Our our free prompt engineering course that we've taught to now more than 7,000 business professionals.
Jordan Wilson [00:36:18]:
Right? That's not really going to work with a new model like o one in a a reasoning model, because it does a lot of this kind of chain of thought thinking, underneath the surface. So right now, this is the only model that, kind of really reasons and behaves like this, but we will I I I I do think that this new update from OpenAI is going to kind of force, Claude and Anthropic to release some updates both from an, a pricing and an API perspective. Right? Because what these models are capable of for other companies using the API, I'm I'm pretty sure anthropic has to be sweating. If I'm anthropic, y'all gotta release something in the next couple of weeks, because I know people are jumping ship. Everyone's either going to open AI with this new o one model or opting for, Google Gemini because their I think their API product is great. I think their front end product is not, the front end Gemini chat, but the the back end for API is great. So pretty significant news here, from OpenAI and, you you know, kind of talking that, hey. This is a pivot at least for our, you know, OpenAI's own internal path toward AGI and saying, hey.
Jordan Wilson [00:37:34]:
We have now achieved a reasoning model, a model that can think under the hood, and now the next, iteration is agents. And 2 things 2 things on that. Right? And just kinda set the record straight because, yeah, this, o one model, a lot of people are talking about it. I think a lot of people are getting it wrong. Well, first of all, it's not a GPT model. I already talked about that. 2 other things that you need to know and that I think are important. Right? Number 1, this isn't the most, powerful o one.
Jordan Wilson [00:38:06]:
Right? This is technically o one preview and o one mini, which is what we all have access to. Right? But at least on initial benchmarks, the actual o one model, which has not been released, is significantly more powerful than what we've all been playing with over the last 2 weeks. Right? Wish the limits were higher. If you're a front end user, you'll only get 50, prompts a week with o one preview, and you get 50 a day with o one mini. So it is very limited, but here's the other thing y'all. And we talked about this last week on our AI agent show. So, I recommend you go back and and listen to that one. That one took a a lot of research, and I think it's really worth listening to episode 361.
Jordan Wilson [00:38:53]:
But right now, o one does not have access to tools. Right? So I think when you combine, right, when we kind of get the quote unquote upgrade from o one preview, and we get the o one model, and then we add the tools that the GPT class has available to the to it. I think right there, that is a very powerful agent. My gosh. Right? Because right now this new model o one does not have access to the web. Right? So it doesn't have access to use the tool browse with Bing. Right? We talked about on our AI agent show kinda what separates a large language model, to, from an an agent, and one of them is tool access. Another one is, access to data.
Jordan Wilson [00:39:38]:
So right now, you can't easily share your data with the o one model. You can't upload files like you can with the GPT class. So all these things that you can do with GPT-4 o. Right? Access the web with browse with Bing. You know, code interpreter or advanced data analysis. You know, browse with Bing, DALL lead, the ability to upload files. Right? The ability to, you know, kind of chat with other GPTs and and keep it in the context window. All of these outside tools, we do not have access to.
Jordan Wilson [00:40:07]:
So we have very limited access to the o one model in a silo. Right? So I I I do think this is going to change, both just how we work and it's going to change what's capable. Right? Imagine when o one, as an example, comes to copilot. Right? So we talked about this on our AI agent show as well. Copilot with their wave 2 announcements for Microsoft 365 Copilot, pretty powerful. Right? But they did say, hey, we will eventually be rolling out this new o one model. So right now, Copilot, in their wave 2 announcements powered by GPT 4 o, but imagine when you get the o one model. This is changing how work is done in front of our very eyes.
Jordan Wilson [00:40:52]:
Alright, y'all. I hope this was helpful. If it was, please let me know. Please share this, share this with your network. Right? It takes us many hours, every single day, to bring you free unbiased high quality information in news about AI, about gen generative AI. Because I want you to be the smartest person in your company when it comes to AI. I want everyone to turn to you and be like, hey. What do we do here on this AI initiative? I want this to be the place everyday AI where you cannot just keep up with what's going on in AI, but where you can get ahead.
Jordan Wilson [00:41:31]:
So if this has helped, please, if you're listening on the live stream, please, share this with your network. It takes a couple of seconds. Tag a friend that needs to to hear this, that needs to know about everyday AI. If you're on the podcast, please leave us a rating, subscribe, and and reach out. You know, check the show notes, reach out, but more importantly, go to your everyday ai.com. Sign up for the free daily newsletter. Every single day, we recap our livestream episodes. So if this was a lot, if your head's spinning, don't worry.
Jordan Wilson [00:41:58]:
We're gonna be breaking down the recap on our newsletter. So thank you for tuning in, and we hope to see you back tomorrow in everyday for more everyday AI. Thanks y'all.