Paul and Mike look into the “tricks-and-treats” of AI news this Halloween on Episode 70 of the Marketing AI Show! This week, our hosts examine the White House Executive Order on AI, Apple’s announcement to spend $1 billion on AI yearly, and OpenAI’s Chief Scientist’s discussion on his hopes and fears for the future of AI.
Listen or watch below—and see below for show notes and the transcript.
This episode is brought to you by our sponsors:
Use BrandOps data to drive unique AI content based on what works in your industry. Many marketers use ChatGPT to create marketing content, but that's just the beginning. BrandOps offers complete views of brand marketing performance across channels. Now you can bring BrandOps data into ChatGPT to answer your toughest marketing questions.
The AI for Agencies Summit is a virtual half-day summit happening on November 2. The AI for Agencies Summit is designed for marketing agency practitioners and leaders who are ready to reinvent what’s possible in their business and embrace smarter technologies to accelerate transformation and value creation. To register, go to AIforAgencies.com and use the code AIPOD50 to get $50 off your ticket.
Watch the Video
00:03:16 — White House issues ambitious executive order on development and oversight of AI
00:25:17 — Apple is set to spend ~$1B/year in developing its own generative AI capabilities.
00:31:58 — Ilya Sutskever discusses hopes, fears, “earth-shattering” effects of AI in the future
00:39:59 — DadJokeGPT, experimenting with generative AI creativity
00:45:20 — MAII’s hands-on experience with Descript’s Overdub + AI Actions
00:48:23 — Amazon rolls out AI-powered image generation
00:51:29 — Generative AI, SEO, and what leaders should know: Google search in the future
00:56:26 — Google agrees to invest up to $2 billion in OpenAI rival Anthropic
President Biden Issues Executive Order on Safe, Secure, and Trustworthy Artificial Intelligence
U.S. President Joe Biden has made big waves in AI policy by signing a sweeping executive order to guide the development and oversight of artificial intelligence in the U.S. The order introduces new consumer protections, requires companies to meet safety standards, and tasks federal agencies with regulating AI risks.
There are a significant amount of parts to this order, but here are a handful that jumped out to us:
The order includes requirements that companies developing AI systems share their safety test results with the U.S. government.
The Department of Commerce will develop guidance for content authentication and watermarking to clearly label AI-generated content.
One of the main intentions emphasizes Americans’ privacy by prioritizing federal support for accelerating the development and use of privacy-preserving techniques—including ones that use cutting-edge AI and that let AI systems be trained while preserving the privacy of the training data.
The order intends to ensure fairness throughout the criminal justice system by developing best practices on the use of AI in sentencing, parole, probation, pretrial release and detention, risk assessments, surveillance, crime forecasting and predictive policing, and forensic analysis.
The briefing addresses AI’s potential to transform education by creating resources to support educators deploying AI-enabled educational tools, such as personalized tutoring in schools.
The White House intends to develop principles and best practices to mitigate the harms and maximize the benefits of AI for workers by addressing job displacement; labor standards; workplace equity, health, and safety; and data collection.
The order also addresses a means to accelerate the rapid hiring of AI professionals as part of a government-wide AI talent surge led by the Office of Personnel Management, U.S. Digital Service, U.S. Digital Corps, and Presidential Innovation Fellowship.
Apple, caught by surprise in generative AI boom, to spend $1 billion per year to catch up
According to a report from Bloomberg, Apple is now spending around $1 billion per year to develop its own generative AI capabilities. This massive investment comes as Apple plays catch-up to competitors like Google, Microsoft, and Amazon, who have already released new AI products and features.
Apple products like your iPhone use AI to do many things. But the company doesn’t have a dedicated generative AI product like ChatGPT yet.
Apple has built its large language model known as Ajax. And there are rumors it has an internal chatbot known as Apple GPT.
According to reporting from Bloomberg, Apple might be looking to incorporate AI into Siri, Messages, and Apple Music to start
A rare interview with Ilya Sutskever, OpenAI’s chief scientist, reveals his hopes and fears for the future of AI
Ilya Sutskever is certainly one of the lesser-known OpenAI cofounders outside of tech circles, but his contributions to the advancements of AI over the last decade are huge—and he just gave a rare interview to MIT Technology Review.
In this wide-ranging discussion, he makes some bold statements about the sheer power of AI today—and in the near future—and speaks of artificial general intelligence (AGI) as a near-term inevitability.
Because of that belief, he’s now working on “superalignment,” or making sure that humans remain in control when AI achieves “superintelligence” (i.e. it’s able to outsmart us at every cognitive task).
When asked about the risks and rewards he sees coming from AI down the line, he says at one point: “It’s going to be monumental, earth-shattering. There will be a before and an after.”
Links Referenced in the Show
- President Biden Issues Executive Order on Safe, Secure, and Trustworthy Artificial Intelligence
- Apple, caught by surprise in generative AI boom, to spend $1 billion per year to catch up
- Exclusive: Ilya Sutskever, OpenAI’s chief scientist, on his hopes and fears for the future of AI
- Paul Roetzer's LinkedIn Post: DadJokeGPT Halloween Costume
- Fix recorded speech as easily as typos with Overdub + AI Actions from Descript
- Generative AI, SEO, and what leaders should know: Google search in the future
- Amazon rolls out AI-powered image generation to help advertisers deliver a better ad experience for customers
- Google agrees to invest up to $2 billion in OpenAI rival Anthropic
Read the Transcription
Disclaimer: This transcription was written by AI, thanks to Descript, and has not been edited for content.
Paul Roetzer: [00:00:00] You don't assume everything is just going to work out the faster you progress technology, but you tend to focus on tangible problems rather than long-term threats.
Paul Roetzer: Welcome to the Marketing AI Show, the podcast that helps your business grow smarter by making artificial intelligence approachable and actionable. You'll hear from top authors, entrepreneurs, researchers, and executives as they share case studies, strategies, and technologies that have the power to transform your business and your career.
Paul Roetzer: My name is Paul Roetzer. I'm the founder of Marketing AI Institute, and I'm your host.
Paul Roetzer: Welcome to episode 70 of the Marketing AI Show. I am your host, Paul Roetzer, along with my co-host, Mike Kaput. And we have A lot to talk about today, don't we, Mike?
Mike Kaput: Yes, we do. Some of it very, very recent.
Paul Roetzer: Yeah. Yeah. Very recent. okay. So it is, it is [00:01:00] 10 30 a.m. Eastern time, Monday, October 30th.
Paul Roetzer: We were recording a little bit later than normal because The White House dropped the fact sheet on the executive order on AI this morning. So we are going to get into that for sure. but first, thank you to our, episode sponsor brand ops. Many marketers use ChatGPT to create marketing content, but that's just the beginning.
Paul Roetzer: We sat down with the brand ops team. We were impressed with their complete views of brand marketing performance across channels. Now you can bring BrandOps data into ChatGPT to answer your toughest marketing questions. Use BrandOps data to drive unique AI content based on what works in your industry.
Paul Roetzer: Visit brandops.io/marketingaishow to learn more. and see BrandOps in action. And this week's episode is also brought to us by the Marketing AI Institute AI for Agency Summit, which [00:02:00] is this Thursday, November 2nd, from 12 to 5 ET. So if you are a marketing agency, or if you hire marketing agencies, I would recommend attending.
Paul Roetzer: It's going to be an amazing virtual event. It will all happen, via Whova is our event platform. there's already, I don't know, I think we're over 220 or so registered. It's a couple hundred agencies already registered. So it's going to cover everything from, I don't know, what is the future of an agency look like with all of the things happening in AI?
Paul Roetzer: that was the big thing for me when I was building my keynote was what is the future state of agencies now? What is the demand going to look like for services? What's the talent going to look like? And so we really built the agenda to try and address all of that. We have, building advisory services around AI.
Paul Roetzer: We have intellectual property. We have. Tools and partner programs. really everything you need to get your agency on the right path heading into 2024, [00:03:00] trying to figure out the impact AI is going to have on your business. So check out AI4Agencies. com Again, that is Thursday, November 2nd, starting at noon Eastern time.
Paul Roetzer: So there is still time to join us. Again, AI4Agencies. com and use the code AIpod50 for 50 off your ticket. Alright, Mike. Let's uh... Let's kick it off. I know we're starting off at the White House. We got a bunch of stuff going on. So it's all yours. All right.
Mike Kaput: Sounds good, Paul. So first up, U. S. President Joe Biden has made some very big waves in AI policy by signing a sweeping executive order to guide the development and oversight of artificial intelligence.
Mike Kaput: in the US. Now, Paul, like you mentioned, we're actually recording this on the morning of Monday, October 30th. And the order came out just a few hours before we started recording today. So this order includes things like new consumer protections around AI. It requires companies to meet [00:04:00] certain safety standards, and it tasks federal agencies with regulating AI risk.
Mike Kaput: Now, there are a ton of different elements to this order. And we're going to unpack some of them. Here are a handful of them across some of these different topic areas that jumped out to me as particularly significant. First up, the order actually includes requirements that companies developing AI systems share their safety test results with the U.
Mike Kaput: S. government in certain cases. The Department of Commerce is going to be developing guidance for content authentication and watermarking to clearly label AI-generated content. we are going to start protecting American's privacy by prioritizing federal support for accelerating the development and use of privacy-preserving techniques.
Mike Kaput: And that can include ones that use AI and let AI systems be trained while preserving the privacy of the training data. There's a stipulation about [00:05:00] ensuring fairness. in the use of AI throughout the criminal justice system. We are also getting work on shaping AI's potential to transform education by creating resources to support educators who are deploying AI-enabled educational tools.
Mike Kaput: There is going to be the development of principles and best practices to mitigate the harms and maximize the benefits of AI for workers by addressing things like job displacement, data collection, and labor standards related to AI. And last but not least, and again, this is just a sample of what's in there, the government is accelerating the rapid hiring of AI professionals as part of a government-wide AI talent surge that's led by the Office of Personnel Management and a couple of other federal organizations.
Mike Kaput: Now, Paul, this statement released by the White House says, quote, with this executive order, the president directs the most sweeping actions ever taken to protect Americans from the potential risks [00:06:00] of AI systems. How big a deal is this actually?
Paul Roetzer: It is a really big deal. so I think it's important to first know that executive order isn't law.
Paul Roetzer: the president can't make law without Congress, but it. It has significant meaning and, they can enforce it in a lot of different ways, directly and indirectly, I guess. I think that first we have to accept this is it. this is what we're going to get in the United States for the time being.
Paul Roetzer: I don't know that they would do any follow on executive orders, but We have to be realistic that the current state of the U. S. Congress is probably not coming to any bipartisan agreements on any new laws around AI. We obviously had enough trouble just getting a speaker, in our Congress. So I just don't see anything major happening before 2025.[00:07:00]
Paul Roetzer: Before we're probably going to get any real laws around this. So I think it's, it's significant in that it has far-reaching implications, but it's also significant in that this is what we're going to probably have in the United States related to laws and regulations. So recurring topic on this podcast is what is the U S doing about this?
Paul Roetzer: This is your answer. This is what the U S is, is doing and is going to be doing probably for the next 16-plus months is my, my best guess at the moment. So for me. we knew this was coming. I saw it. I think I put it on LinkedIn on maybe Saturday night. There was a political article that sort of tipped off that this was likely coming on Monday morning.
Paul Roetzer: So you and I were ready for it. We're hoping it was going to drop before this podcast started. and it did but I would also say as of 10 30 a.m. Eastern Time Monday morning. All we have is the fact sheet right now. So the White House released a fact sheet and they launched AI. gov or maybe AI.
Paul Roetzer: gov was a thing and it was the first time I went on it this morning. [00:08:00] We do not have the full executive order yet, so the commentary we can provide is based on the fact sheet and what we've been able to ascertain since then. So this morning, I started by going back and looking in October 2022, the U.
Paul Roetzer: S. government released the blueprint for an AI Bill of Rights. And in that blueprint, they outlined five principles around the design, use, and deployment of automated systems. to protect the American public in the age of artificial intelligence. And so I think that blueprint sort of started, started to set the stage for the things we should expect to see in this executive order.
Paul Roetzer: And at that time they dealt with safe and effective systems, discrimination and protections for consumers, data privacy. notice and explanation when automated systems were being used. Again, it's largely about the protection of consumer rights and then human alternatives, considerations, and fallback, meaning you should be able to opt out where appropriate and have access to a person rather than AI if you want it, [00:09:00] things like that.
Paul Roetzer: So we, we had some indications. Um. What I was looking for, what , I made my list before I read the fact sheet about what I'm, what I want to see in it or what I expected. The first was jobs in the economy. The second was misinformation and disinformation. The third was elections and democracy.
Paul Roetzer: The fourth was consumer safety. The fifth was impact on open source AI and the concept of regulatory capture that the big tech is basically driving these things. So that's some context around how I viewed this. The other thing I think is really, really important, because we're going to get into a little bit here of our perception of this executive order, or at least the early iteration of it we're seeing.
Paul Roetzer: But I think as you try and form your own perspective on this, you have to understand Who you're listening to, and this, this goes for anything. This goes for politics. It goes for media. You have to understand the [00:10:00] context of why people say what they say. So the way I look at this, and I started outlined this morning of the different groups who would talk about this, the first one you have to be very conscious of are the doomers.
Paul Roetzer: These are the people that are winning most of the mainstream media headlines right now about AI and the existential risk of humanity. They would be considered the loud minority. So if you look at the AI research community, there are very few of them who actually worry intently about existential risks to humanity from AI.
Paul Roetzer: But the ones who do, Geoff Hinton, are the ones who are on 60 Minutes touting This stuff and worrying about it. So they believe generally. So when you again, you have to assess the lens. Someone is looking at this executive order through. So the doomers believe we should slow down. They struggle to find practical explanations of how they arrive at these doomsday scenarios.[00:11:00]
Paul Roetzer: But they generally lean on this idea that AI is going to become super intelligent, or it's going to emerge somehow, and that the AI is going to have no use for us humans or bad actors are going to exploit this technology in some incredibly nefarious way that leads to a catastrophic outcome. But they, they struggle often to make these, these arguments tangible.
Paul Roetzer: The thing that's tricky about the Doomers is their motivations seem very honest. there, there doesn't seem to be ulterior motives to someone like a Jeff Hinton. why do you leave Google and go out on a media tour, which you've never done and talk about these things. it appears as though he truly believes this.
Paul Roetzer: So that's the Doomer mentality, but most AI researchers Believe the doomers are misguided. So again, it is the loud minority. It is not the common viewpoint. So you may hear doomers relation. The other one is that we talked about last week is the optimist or the accelerationist. [00:12:00] These are the people that believe that rapidly advancing technology leads to growth and abundance, period.
Paul Roetzer: That there are downsides, there are negatives, but at the end of the day, the faster you move, the more you create technology, the better off humanity is. So to go back to Marc Andreessen's Techno Optimist Manifesto and pull two quick excerpts. Technology is the glory of human ambition and achievement, the spearhead of progress, and the realization of our potential.
Paul Roetzer: We believe growth is progress, leading to vitality, expansion of life, increasing knowledge. Higher wellbeing. So the optimist accelerationist will have a different perspective on the executive order. The third, I just generically called big tech. I didn't have a, I don't know what else to call these people.
Paul Roetzer: They are the ones pushing for the regulation. This is Google, Microsoft, OpenAI, anthropic inflection, all the leaders in the foundation model business right now, the ones building the most powerful models. The [00:13:00] ones that are loudest about the need for regulation. Now these are tricky ones to assess because on the surface, it seems altruistic, it seems they truly believe in the importance of regulation, but in reality, , they have significant financial motivations to maintain the lead that they've established, so they have the most to gain, and they also have The most to lose.
Paul Roetzer: So it's hard to assess the true meaning of why big tech is pushing for this. And then the fourth group, which is where I would put myself. I think Mike, this is probably where you would follow. I don't want to speak for you, but you can, you can offer your thoughts is the rationalists. And so this is the people who are focused on near term challenges that are presented by the current models and the ones we can project out, for the next year.
Paul Roetzer: We're worried about autonomous weapons, systematic bias and products and services, disinformation and misinformation, synthetic media, disruption to [00:14:00] jobs, educational systems and the economy. The rationalist can be pro AI, you can be for progress and technology, and you can believe some of the things that the optimists believe.
Paul Roetzer: But you try and balance that with human interests and ethical considerations. So we don't in the rationalist world, you don't assume everything is just going to work out the faster you progress technology, but you tend to focus on tangible problems rather than long term threats. So again, as we've said on the show, like.
Paul Roetzer: I don't not worry about existential risk to humanity. I don't, I'm not sitting here saying there's a zero chance that something catastrophic happens, but I don't find it productive to spend my time thinking about that very much. So I'll stop there for a minute, Mike, and just see, to make sure. See if there's any thoughts you have, maybe on the four personas, cause you and I haven't even talked about this.
Paul Roetzer: This was literally this morning. I'm trying to wrap my head around how to assess this and how to share this with people. so did you have any [00:15:00] thoughts on that persona approach? Yeah.
Mike Kaput: A couple of things. One, I love the approach. I actually think it's probably a pretty smart way to approach any subject.
Mike Kaput: This is a life. Pro tip for sure. It's understanding the landscape of interested parties and the broad groupings of them. And I would also agree with those groupings. And I think in some cases, probably what's important to reiterate to people is yes, everyone in these groups has different agendas, but you would probably be surprised how many people in big tech companies are both incentivized with that financial motivation, which just happens Is reality, but also how many of them seem to truly believe that the techno-optimist outcomes are the ones that they.
Mike Kaput: Should be working on and working on in that way. So this whole, the Silicon Valley Phrase move fast and break things, right? I think was maybe Coined by one of the one of the big tech companies. [00:16:00] Maybe it was Facebook. I don't recall but basically made popular by Facebook Yes, and it's this mentality that If you that breaking things is not only okay, but sometimes it's advisable because it means you're moving fast enough to race to market to develop certain technologies.
Mike Kaput: And that is exactly what we're seeing here. And I think the rationalist perspective is probably Hey, let's move fast and break fewer things, right? Or at least be concerned about how they can break. Yeah. So yeah, I would agree generally with your assessment. I think it's spot on and it's really, really important for people to understand these different groups.
Mike Kaput: And I think some of it sounds like from some of the reporting so far on the, on the executive order, a lot of this came out of the government looking at their perceived failure with social media as a technology. I think some people within the Biden administration, it sounds like felt like they had dropped the government [00:17:00] had dropped the ball.
Mike Kaput: On stopping these people from breaking things by moving so fast. And so you're seeing this, but just on steroids. So they really, I think they recognize that, which is positive.
Paul Roetzer: Yep. Yeah. And I think so going back to then how I, the lens with which I look at the executive order, I look at personally as a rationalist, that is just how I generally approach everything in life is trying to be reasonable and logical about the arguments and trying to think of both sides.
Paul Roetzer: I think you have to accept that the executive order is most likely largely influenced by big tech. Of the four groups that is largely who has been sitting in these meetings It appears they've tried to balance it with some of the doomers. There's certainly those voices have had an effect in different ways I think they've pulled in some of the optimists Accelerationalists some of the the open source advocates the people who are really championing [00:18:00] the need for open source but Most likely big tech would be the most influential group in this, this persona analysis.
Paul Roetzer: So with all that being said, here's my overall take so far. Again, keeping in mind all we have is a fact sheet and an outline of key things, the fact sheet does not offer very many specifics, if any specifics on how going to do it, how fast they will do it, or how much money the government will invest to do it.
Paul Roetzer: So while there are lots of really promising ideas here, and they do address a wide range of issues. There isn't much in this fact sheet that tells you when any of this will happen or how actually going to adhere to it, the idea of, AI generated content being, tracked and reported.
Paul Roetzer: How? that may be a five year thing before they have any way to do that with [00:19:00] text. we know images that might be possible. And When I, when I go back to my list of the things I was looking for jobs in the economy, basically said we're going to study it, that there will be a study conducted on the impact of labor and that we will make it easier for people to come to the United States and work on this technology.
Paul Roetzer: And in fact, AI dot gov, the navigation of that site. Has build your AI skills, which links to a bunch of foundations and stuff. And then it has bring your AI skills to the U S those are literally two of the four links in the navigation. So that was my take on jobs in the company, misinformation and disinformation.
Paul Roetzer: They certainly address it. and they largely talk about the need to identify the AI generated content, which again, there's no near term solution for, don't know when going to do that or how going to do it. They don't directly talk about elections and democracy. It's sort of probably passively related to a lot of things that are mentioned.
Paul Roetzer: Consumer safety, it's all over it. That's basically the premise. It's the thing they can control because [00:20:00] they can't create laws, but they can force government agencies to enact or to, govern existing laws. More aggressively and they can push for agencies to adhere to new policies and principles so they can affect consumer safety probably the most and then the impact on open source and this idea of regulatory capture.
Paul Roetzer: It's honestly not clear to me yet, I need to see more commentary. From those people, to see how perceiving it. Cause it was not, it was not very, clear to me how they would. So the two people I have seen, as I mentioned, I have not seen much commentary at all yet this morning. And this is three and a half hours old.
Paul Roetzer: One was Jack Clark, who is the co founder of Anthropic. I would put them in the big tech category. He was also previously policy director at OpenAI. And an AI reporter at Bloomberg prior to that, he had about a six tweet thread, but one said proof is obviously going to be in the implementation, but the executive order gets a bunch of different parts of government to create capacity [00:21:00] for third party measurement, oversight and analysis of AI systems, which seems broadly helpful.
Paul Roetzer: And then we had Aaron Levy, the CEO of Box. He said Biden's AI executive order is the gold standard for how governments should be regulating AI right now. Thoughtful but scoped oversight focused on practical risks, emphasis on privacy and security, focus on R& D across the ecosystem, and encouraging use in, of AI in the government.
Paul Roetzer: That would seem a rationalist take to me. and again, I lean in the direction of I think this is an accurate assessment of it, but until I read the actual 111 page or so executive order, I can't say this to be true because there just isn't enough information to go on yet.
Mike Kaput: So it sounds as some of the initial reporting from the Associated Press around this, which definitely had some quotes directly from people within the administration, it sounds like they've been thinking about this issue for a while but also planning [00:22:00] potentially this to be a springboard for broader, broader voluntary commitments from tech companies, but also possible legislation.
Mike Kaput: But what you're saying here is it sounds like this is what we're getting for the next year and a half. So people should be considering not holding their breath probably for some sweeping legislation like the AI act in the EU. Is that correct?
Paul Roetzer: I hope I'm pleasantly surprised by Congress's ability to figure out how to do anything at all right now, more or less be part of, be bipartisan on an issue.
Paul Roetzer: so I hope that happens, but no, I wouldn't hold my breath thinking that Congress is going to somehow come together and put laws in place, but I would love to be wrong.
Mike Kaput: And so that's not totally surprising. The regulatory framework, or at least the way it looks like it's shaping up, is going to be very, very different in the U.
Mike Kaput: S. compared to the E. U., which has been broadly [00:23:00] true with a lot of data privacy and technology focused legislation over the last several years.
Paul Roetzer: Yeah, and it's going to be one of those things where the comp keep in mind that the companies building these foundational models are almost Exclusively based in the US, at least the ones that are going to be broadly used right now.
Paul Roetzer: And so going to be affected by EU's AI Act regardless. And it may impact the way that they have to build these things anyway. So it's not the EU isn't going to have any effect on the US. It could, but I just feel like. It's going to take 2024 to see a lot of these ideas through that are even outlined in this fact sheet and to figure out how to govern these things and to get the studies underway that they talk about because so much of the fact sheet of the executive order is about.
Paul Roetzer: Research and analysis and establishing frameworks and putting bodies in place to do this and hiring AI researchers within, government agencies. And it's, you don't flip a switch and [00:24:00] see all these ideas come to life at January, 2024. This is going to be a prolonged implementation of this vision, I guess.
Mike Kaput: Yeah, and as we wrap this up here, just a quick weird shout out from this article, they go into all the extensive steps that by the administration and gone through to understand artificial intelligence, talking with experts doing a lot of research and fact finding. So it's very clear.
Mike Kaput: , they were deeply involved in the subject, but they had stuff bizarre shout out to the new mission impossible movie. They spent a paragraph talking about how, as all this stuff was happening and being formulated that Biden watched the new mission impossible movie, which apparently is about AI taking over.
Mike Kaput: It's the villain. And came away with a better appreciation of what could go wrong. And I was I don't know if that's going to be your primary source for this, but Hey, shout out to mission impossible for raising awareness of AI risks.
Paul Roetzer: For some reason, I'm picturing like a weekend of movie night at the [00:25:00] white house where watching like "Her" and "X Machina," "Mission Impossible," on like.
Paul Roetzer: Now I'm a little worried.
Mike Kaput: It was not, I want to make it very clear. That's not the only thing that made them aware of this, but it was a funny and bizarre shout out. All right. Our next main topic today, according to a report from Bloomberg, Apple is now spending about a billion dollars per year generative AI capabilities.
Mike Kaput: So this investment comes as Apple is playing catch up. To all the other big tech companies like Google, Microsoft, and Amazon, all of whom have released a bunch of AI products and features. Now, Apple products make no mistake. They use AI a lot, like your iPhone relies on many types of artificial intelligence to do what it does, but they don't have a dedicated generative AI product like ChatGPT.
Mike Kaput: Just yet now there have been past reports that they have built their own large language model Which is known as Ajax and [00:26:00] there are rumors that Apple has an internal chat bot known as Apple GPT According to Bloomberg Apple might start looking to incorporate AI into Siri messages and Apple music to start.
Mike Kaput: So Paul, I know you follow Apple very, very closely. Can you give us a sense of how Apple has approached AI up until this moment?
Paul Roetzer: Yeah, they don't talk about it. so Apple is extremely advanced with AI. They build their own chips. They have, probably hundreds of AI features and iPhones and watch and the iPad and max, and it's an AI company.
Paul Roetzer: They just, for whatever reason, historically never talk about it. so that's first and foremost, they make massive investments. Contextually though, I think it's important because it's like always you have to understand the broader implications. A billion dollars sounds like a lot of money to people that is less.
Paul Roetzer: It's probably like 3 [00:27:00] percent of their R&D budget. If I'm doing my math in my head correctly, I think in 2002, they spent over 26 billion on R&D. So my initial feeling is a billion is not enough now because of how they classify things. The other 26 billion likely has all kinds of AI tied to it.
Paul Roetzer: Like when building products, they don't build them without doing machine learning and deep learning. but they just don't classify it that way. I don't know. Contextually, my first reaction was a billion is nothing to Apple. the other thing I think about though is I wonder if they, they aren't forced to change their messaging soon.
Paul Roetzer: So I'm actually curious because they have an event tonight. So again, this is October 30th. We're recording this. There's an event tonight called. Scary fast is the theme of the event. and I think it's going to be related to chips, there's silicon chips, and the processing speed and the efficiency, energy efficiency and things like that, but there's [00:28:00] rumors.
Paul Roetzer: There might be some other stuff that's talked about. It's an unusual time. They don't usually do 5 PM I think it's a, yeah, 8 PM Eastern time, 5 PM Pacific time event, which is highly unusual for Apple. So again, maybe they'll do something, but historically they just. Embed it within and they don't even tell you it's there.
Paul Roetzer: So again, chips. One of the great examples to me is Photos you know You can go into apple photos on your phone and search for any image or any time or place or person And it just shows up magically shows up Well, they've been using ai for years to enable that and they don't even talk about it versus google which is running ads right now showing how you can take a series of photos and it'll magically pick the best face, the best smile of each person in the photo and replace them all into a single photo.
Paul Roetzer: So it's like stitching the photo together using AI and they tell you it's using AI. So Apple is a very stealthy AI powerhouse. They don't publish their [00:29:00] research. They don't have their people rarely have public personas. So they don't have people who are out, active on Twitter. Apple itself is barely active on Twitter.
Paul Roetzer: It's they just quietly go about their business building smarter products and they just work and they seem like magic. I don't know that they can do that moving forward. I keep waiting for them to all of a sudden start talking about AI and I don't know. I'm interested to see if it happens this fall.
Mike Kaput: So talking about generative AI capabilities and products specifically, can Apple catch up at this point? The space moves so fast and the other players have made such huge moves. I'm wondering how you see how far behind they are,
Paul Roetzer: if at all. Well, they definitely I mean the article you referring to said they got caught flat footed they which is again I'm just still shocked by how How unprepared Google and Apple and other these other companies were for the [00:30:00] ChatGPT moment.
Paul Roetzer: it's, it's really shocking to me. I still can't quite process it. Knowing all the reason the AI researchers they had and access to this stuff that they had. but, I think that was, my initial take on where they were. And then I think where they go is I don't know that anything but Siri matters, like going to infuse it into, I assume into keynote, we use keynote, you and I are both fans of Apple keynote.
Paul Roetzer: We use that product. you could see it put, put into a bunch of different products, but if they don't make Siri smarter. It's just one of the great misses in business history. Cause even we'll talk about in a couple of minutes, my experience using the voice conversation in chat, CPT over the weekend.
Paul Roetzer: And it's gotten pretty good, to have a conversation, you can do the same thing with inflection pie. you can just talk to the AI. Well, that's what Siri was supposed to be. So again, I, all the stuff that Apple [00:31:00] does, unless they make Siri. What it was supposed to be. It's just a massive miss on their part.
Paul Roetzer: And so I got a guess that working on that.
Mike Kaput: So Siri is really the game changing opportunity that they have when it comes to actually doing something very meaningful publicly with Gen AI.
Paul Roetzer: Well, yeah, because it's infused into everything. So imagine being able to talk to your computer and say, open up this application, do, take this thing, do this action, talk to your computer.
Paul Roetzer: The same thing with your phone, not just again getting weather and scores and your calendar for the day, but having a conversation with it, finding out, having it do tasks for you that you don't have to set the rules for. right now you can do things in your iPhone, but you have to set all the rules yourself.
Paul Roetzer: I just have that conversation, into your iPad, your watch, the vision pro, every product they have to be made so much better from a consumer experience standpoint, if you could just talk to them [00:32:00] and they know that, that's what I'm saying. there's no way that isn't the thing that trying to solve.
Mike Kaput: So in our third big topic today, we're going to spend a minute talking about a guy named Ilya Sutskever, and he is one of the lesser known co founders of OpenAI, and he is their current chief scientist. he, despite being lesser known, has made huge contributions to the advancements of AI, both at OpenAI and before he was there, working originally with Jeff Hinton to develop some of the core technologies that are making generative AI today possible.
Mike Kaput: And he actually just gave basically a rare interview to MIT Technology Review. It's a wide ranging discussion and well worth reading all of it. He makes some really bold statements about the sheer power of AI today and how powerful it will be in the near future. And he speaks of Artificial General Intelligence, or AGI, as a near [00:33:00] term inevitability.
Mike Kaput: And because of that belief, he's working on something calling Super Alignment at OpenAI, which is making sure that humans remain in control when AI achieves super intelligence, which They would define as it's able to outsmart us at every cognitive task. Now, when he was asked about the risks and rewards that he sees coming from AI down the line, potentially very soon, he says, quote, it's going to be monumental earth shattering.
Mike Kaput: There will be a before and an after. Now, Paul, walk us through why this interview is so important and maybe why it matters, even though we're talking about some very big picture topics here.
Paul Roetzer: I think first, just his significance to the moment we're in an AI. Most people don't know, again, outside of tech circles, the average person's probably never heard of Ilya, unless you've read Genius Makers by Cade [00:34:00] Matz, which is where I really got introduced to him as a significant player.
Paul Roetzer: in this current age, so I think a number of things, we always talk about the context of understanding the mentality of the people behind this technology, the people who are building this and what they believe, you don't have to agree with it, but this is what they believe to be true. So the AGI thing is a big thing like.
Paul Roetzer: It's becoming more, common to talk about this year. I remember last fall sitting there thinking I need to do a, a LinkedIn post about AGI, but I don't know that people in my network are really ready for that conversation, but it's important to me that they start to realize this is a significant thing.
Paul Roetzer: I don't remember if I ended up doing the post or not, but I remember that moment of thinking people need to be aware, but I don't know that ready yet. And so to hear them talking about it as this inevitability, he truly moves back and forth between [00:35:00] present tense and future tense.
Paul Roetzer: When he talks about a AGIt's almost it's already there, which if you believe some of the rumors, it may be opening. may think that they actually have AGI and they just haven't told the rest of the world about it yet. but they talk about it as inevitable, which was not a common belief even, a year ago.
Paul Roetzer: the before and after, the monumental earth shattering thing you mentioned, that's interesting because again, to many people, what we have right now is earth shattering. Talking about what we have right now is... Just boring AI that isn't AGI. And so to them, there is this before and after world.
Paul Roetzer: I think the article is really cool if you appreciate the history of this stuff because it does go into AlexNet, which they called the big bang moment in 2012 for deep learning, the rise of NVIDIA GPUs, the significance of Ilya and how OpenAI was founded and how his recruitment actually led to them being able to get other top AI [00:36:00] researchers.
Paul Roetzer: One of the stories I thought was really cool is we've heard all these , I guess, related stories about opening eyes, people not believing in ChatGPT that when they released it, they weren't sure it was going to work. And I thought that was really cool because he said. at the time opening, I had no idea what it was putting out.
Paul Roetzer: Expectations inside the company couldn't have been lower. Says Sutskeva, I will admit to my slight embarrassment, I don't know if I should, but what the hell, it is true. When we made ChatGPT, I didn't know if it was any good. When you asked it a factual question, it gave you a wrong answer. I thought it was going to be so unimpressive that people would say, why are you doing this?
Paul Roetzer: This is so boring. And so just to like... Get that context of, of how they viewed things is, is so intriguing. And then, obviously he goes into the AGI stuff and he talks, about how the shift in mentality, even with AI research community, it was thought of as crazy to talk about AGI before ChatGPT, but [00:37:00] they felt like.
Paul Roetzer: that changed. and then, just getting into the whole thing at the end about, humans choosing to merge with AI again, nuts, the average person reading that would never happen. Yeah. And he admits that he's it sounds crazy, but there's probably going to be a time when, when super intelligence exists where it's like.
Paul Roetzer: I'll just join up with the AI because I got, I got nothing as a, as a lowly human. And it's like that thinking about these things is, is just, it's important perspective to dip your mind into a little bit, realize what's going on and then go back to your daily life trying to figure out how to, write better articles with AI.
Mike Kaput: So one thing jumped out to me about it when he was talking about how ChatGPT basically changed everything. And it was not just about getting, more than 100 million users very, very quickly. It was also because quote, ChatGPT [00:38:00] has allowed machine learning researchers to dream. And by that, It sounds a little pie in the sky, but it's actually, I think, really important to understand that this unlocked a lot of very talented AI people to start believing something artificial general intelligence could be possible and sooner than we think.
Mike Kaput: Can you maybe just unpack a little more some of the views around AGI and why these might be more than just science fiction.
Paul Roetzer: Yeah, so we we actually We've talked about this a little bit. We Vedant Mishra was a google researcher who actually did A talk for us at Macon and I remember asking him This question about why are you at Google working on this?
Paul Roetzer: And he said that we, I believe in the possibility of AGI basically. And I said, well, why, it could go wrong. Why would you be working on AGI? And he said, because people like me believe in a world of [00:39:00] abundance that we think that if we achieve AGI, that we can create. An amazing future for humanity. And I don't, again, I don't know how universal that belief is, but I've talked with enough people in the iResearch community who share that vision, that if we can do this, we can create an amazing, abundant future for everyone.
Paul Roetzer: And, and it seems to be a driving force behind why people do this.
Mike Kaput: Yeah, so it really does seem whether we get there or not, and some people don't even think we can, that some people do see tangible work doing at some of these big tech companies that you mentioned as one of the major groups here, that they could actually be ushering in some type of machine intelligence that is broadly as good or better than human beings.
Paul Roetzer: Yeah, which is terrifying and exciting, depending on [00:40:00] how you feel that day.
Mike Kaput: Alright, let's dive into some rapid fire topics, and the first one is going to be something a little lighter, but actually very important. Paul, you actually posted about an experiment you did this past weekend, having ChatGPT come up with a Halloween costume for you and helping you execute it in a number of ways.
Mike Kaput: And not only is that a fun, I think, example of what's possible, but it also... Shows, how you can be creative and change how you plan out basically any project when it comes to with AI, when it comes to basically anything you can think of doing. So could you walk us through what you did and what Halloween costume you ended up getting out of it?
Paul Roetzer: Yeah. So it was interesting. It was a, it was Saturday morning and we were just striking out on Halloween costume ideas and we had a party. At three o'clock that day, a family party. And so I was sitting there with my daughter, my wife, my son was playing video games, I think. And so I said, it'd be funny.
Paul Roetzer: It's if I just [00:41:00] figured out a way to have my phone with me and then someone could come up and push a button on my shirt and ask me anything. And I would just use ChatGPT's voice capability to have a conversation with people and it'd be interesting and educational. And my daughter just like.
Paul Roetzer: Whatever. And I said, no, try it. And so I pulled it up, ChatGPT plus, and I did the voice conversation. I said, just, let's just talk to it. And she just wanted nothing to do with it. And so I said, I need some ideas for Halloween or something. And it came back and said, Halloween's fun and creative.
Paul Roetzer: Are you looking for costume ideas, decoration ideas, or perhaps spooky recipes? Let me know what you're interested in. I'll be happy to help. So I said costume ideas for an adult male, age 45, something fun, something that my kids might find cringy because Gen Z, which of course got the eye roll from my daughter.
Paul Roetzer: because I'm not allowed to use the word cringe, only they are apparently. So one of the ideas that came back with right away was a dad joke detective. And I was Oh, well, I don't. I don't like the detective thing, but I like the dad joke thing. that's pretty [00:42:00] cool. I know chat, you can tell jokes.
Paul Roetzer: So I started prompting it saying, can you create a series of Halloween theme dad jokes? And it creates like seven to eight at a time. And I just kept like more, more, I just kept prompting it. give me more. Cause they were pretty funny. And so I'm reading them and I'm laughing. my wife's , at least giving me a token giggle.
Paul Roetzer: My daughter is just rolling her eyes at me still. So I was this is good. if I'm getting the eye roll, it means I'm onto something. Once I had the dad jokes locked in, I needed a way to deliver this. I needed to know what the costume was going to be. So I said, okay, what are some outfit ideas related to this?
Paul Roetzer: I want to let people come up and request a joke. I'm thinking a robot or Silicon Valley bro, or dressing up as ChatGPT in some way. It came back with a few interesting things. And then I was oh, I wonder what ChatGPT would look if it was a middle aged dad. And so I prompted it. I said, well, what would you actually look like if it were a middle aged dad?
Paul Roetzer: And it came back with two options and one it said, ChatGPT personified as a middle aged dad. The first image portrays ChatGPT as a friendly middle aged dad wearing glasses, a [00:43:00] polo shirt, khaki pants. I was I got all those things. and and he's holding a smartphone, displaying a chat bubble and you can spot a dad joke book peeking out of his backpack.
Paul Roetzer: I was Oh, this is perfect. I can make a logo. I can put it on my iPad. I'll walk around telling dad jokes. this is awesome. So I'm okay, now I need a logo. So now, and again, flip this to a business scenario where you're just. Brainstorming. You're just trying to create ideas where I was nothing.
Paul Roetzer: I was staring at a blank page. So I go through iterations on a logo. It keeps spelling it wrong. It was driving me nuts. it gets, if anybody's used DALL-E 3 and tried to get words into it, to do things like this, it'll just leave letters out randomly. It'll make it cat GPT instead. it does weird stuff.
Paul Roetzer: So I finally got it to work. I said, okay, let's try one more time from scratch. Here's what I need. Logo for product named Dad Joke GPT, which is the name I'd come up with. Halloween themed, full color, fun for kids, please spell out the full name in the logo. And this time it nailed it. And we'll put the, we'll put the images in the, in the show notes on the site.
Paul Roetzer: so if you just go to Marketing [00:44:00] Institute and check out the podcast or the blog posts, you can see these. Or you can go to my LinkedIn page and check them out. And so I said, great work. You nailed it. And ChatGPT said, thank you. I'm glad to hear it. If you have any more requests or need further assistance, feel free to ask.
Paul Roetzer: Enjoy your Halloween themed Dad Joke GPT. Dad joke, GPT logo and have a spooktacular day. And so I said the moral of the story here is. Just play with these tools. this gave me an excuse to play around as a logo generator of I'm wanting to see, can it do it? And it actually did some awesome logos.
Paul Roetzer: that just had the misspelling wrong. if I could have just said, no, change the spelling to this, but use that driving image, which I probably could have done, I guess. I probably could have uploaded that. I don't know. But the whole idea is you can discover ways to use these tools that are so far beyond writing emails and articles and proposals if you just play with them and learn what capable of doing.
Paul Roetzer: And so much of what I just explained is all relatively new capabilities since the GPT 4 vision capability, and then the ability to have a voice conversation [00:45:00] emerged. So it was just a fun, tomorrow's Halloween. It's like. It's a little more lighthearted based on superintelligent stuff.
Paul Roetzer: yeah, we thought it'd be a fun thing to just share with people, but really it's more about this idea of experimentation and competency. And these tools comes from finding use cases, whether it's in your personal life or professional life, you gotta just try them and see what capable of.
Mike Kaput: So for our next rapid fire topic, we're talking again about Descript, which is a popular AI video and audio editing tool that we rely on here at the Institute. And we're going to cover two quick topics related to Descript. Now, first, the company just launched something called AI Actions. These are a series of ongoing AI updates to Descript's capabilities.
Mike Kaput: They say that it's built right into your workflows. Descript's AI tools do the work so you can skip to the flow. Now, right now, these tools, and there are others coming soon, they say they [00:46:00] include things like a chapter generator. So this will give you instant chapter markers and titles. They have something called summarizer that will summarize what is being said in a transcript.
Mike Kaput: They have a social post writer and an ask AI anything feature that can help you brainstorm some ideas. So those are pretty cool and worth checking out if you're a descript user. And second, we wanted to share just a quick hands on experiment we've been doing with Descript's Overdub feature. So this feature allows you to create a realistic digital voice clone of a speaker.
Mike Kaput: And you can use this to dub over things they said in a recording, so they don't have to re record short bits of content. One of our team members said, quote, how easy it was, creeped me out, but also was really cool. So Descript basically makes you record yourself talking or saying a specific paragraph of text that it gives you in order to create this voice clone.
Mike Kaput: Or you can upload someone reading. [00:47:00] The specific paragraph that it requires and once you do that, you then just highlight the text that you want to replace in your transcript, you right click select overdub, and you literally just type in what you want the voice clone to substitute instead so Paul, our team's been having fun with this pretty crazy.
Mike Kaput: Technology. Obviously, we're doing it all with the permission of the people whose voices are being cloned, which is important. What did you make of some of the experiments and some of the new Descript features that we're seeing?
Paul Roetzer: Yeah, we're obviously big fans of Descript. We talk about them a lot.
Paul Roetzer: Not a sponsor of ours, so it is just literally because it's an awesome tool. And we mentioned, I think, a week or two ago that they had a bunch of stuff coming, and we're already starting to see the flood of this stuff. The overdub is awesome. It is definitely creepy how easy it is to do it. I think, what did you say?
Paul Roetzer: It was a 30 second script she read or something and just cloned it. it seems like we're on a bit of an honor system, which is really [00:48:00] creepy, I guess that, anybody can take any audio file and upload it and do it. It seems through basically an honor system. But yeah, the tech is awesome.
Paul Roetzer: When we look at the positive uses of it, which is what most people are going to do with it. Again, most uses of AI will be for good, for creativity, and to augment what we're capable of and enhance our abilities. And in that sense, if we stay on that focus, it's amazing stuff.
Mike Kaput: Next up, Amazon has made some waves in the advertising world this past week with the launch of a new AI tool.
Mike Kaput: So they actually rolled out an image generation capability. As part of Amazon ads, that's in beta, and this basically just helps you generate more visually compelling ads using AI. the solution lets brands generate lifestyle images, which showcase their products being used in context, which Amazon says can boost ad performance.
Mike Kaput: an example [00:49:00] given by the company is an advertiser may have, say, standalone images of their product against, a white background. say a toaster. And when the same toaster is placed in a lifestyle context, being used on a kitchen counter next to a piece of toast in a sponsored brand's ad, the click through rates apparently can be up to 40 percent higher than to ads with just boring standard product images.
Mike Kaput: So this tool will actually help you simulate lifestyle factors around a product without having to essentially. Do a photo shoot and Amazon has rolled this out because it is directly proven that dramatically increases the use or the effectiveness of the ads you're creating. Paul, what did you make of the ability to now do this right within Amazon ads?
Paul Roetzer: Then we're going to see a flood of these kinds of features in existing platforms. The thing that's interesting about this that maybe is a prelude to what we'll see more of in 2024 is. You could use [00:50:00] a mid journey or Adobe or something like that to do this same thing. Just take an image of a toaster and do an image fill or create an image around it.
Paul Roetzer: Create a scene around it. you could do this. But the key is when you do it in Amazon, they can connect it to performance data. And they can, in theory, optimize those images based on what they know works best. And so it's this idea of generative AI connected to analytics. Which is a total on lock moving into next year that we're not seeing much of right now.
Paul Roetzer: when you're in ChatGPT and you're writing emails, it's not connected to the best performing emails. it doesn't know what's worked previously. So I think for existing platforms, whether CRMs or e commerce stores. To enable generative AI that also ties to performance data so that the outputs of the generative AI are more intelligent.
Paul Roetzer: That's a really interesting thing looking forward to next year. And I have not seen very many applications of that. you certainly say with Facebook ads, I assume [00:51:00] with Google ads, you're seeing it in the ad space, but I could imagine seeing that played out in a lot more marketing sales and service as well.
Mike Kaput: Yeah, that's really interesting, especially just unlocking this capability that yeah, you technically could have had before if you had sat down and done new photo shoots or done an extensive amount of editing. This is a really tangible use case that directly leads to better performing ads, which I found great.
Paul Roetzer: Well, especially since a lot of the people that are selling on there are probably small businesses, independence. So yeah, it's a bringing that capability down. To the smaller businesses is a big thing.
Mike Kaput: So in our next topic, a new analysis from our friends over at insight partners. it's basically a big wake up call for every marketing professional, in a recently published article insight partners breaks down how AI specifically Google's AI powered search generative experience SGE will have a disruptive impact on traffic.
Mike Kaput: from search. [00:52:00] Specifically, they say that this conversational nature of the SGE results is going to result in more quote zero click results where basically users get a full answer or a result without having to click on links. we've already seen this playing out as we're experimenting with SGE, and they actually also point out that only 57%.
Mike Kaput: of SGE results are from the first page of organic results, and that the first SGE link is the same as the first organic link, only 12 percent of the time. So their conclusion here is that if users start switching their reading habits and clicking behavior to the SGE box that appears over traditional search results, This could be a huge shift in traffic and eyeballs.
Mike Kaput: And so as a result of this, they make some really solid recommendations to companies who rely on search traffic to power their marketing engine. Notably, they said traffic fuels so many aspects of your go to market engine. [00:53:00] Lead quantity could go down even if quality goes up. Retargeting campaigns may be able to target only a fraction of their previous audience.
Mike Kaput: Lead scoring accuracy may be reduced if more research is done off site. Brand scorecards and awareness campaign measurement could be less accurate. As a result, insight says things like your content strategy needs to change and evolve. They actually expect traffic declines up to 25 percent due to this.
Mike Kaput: And really sounding the alarm to start preparing for 2024, because that's when SGE and similar types of AI powered search results are rolling out more in a more widespread fashion. And basically they conclude. You want to establish what they call a war room for SEO and SEM and figure out Where, where you're going when it comes to search and SEO in your business.
Mike Kaput: So Paul, what jumped out to you here about their advice and their alarm bells that ringing about [00:54:00] search?
Paul Roetzer: We've been looking for some tangible data on this. I think there's a lot of assumptions being made about the potential decline in organic traffic, the impact on SEO, and then downstream, the impact on just content.
Paul Roetzer: Marketing and brand publishing overall. And obviously if you're a media company, the impact on traffic to the site and ad clicks and things like that. So there's just a lot of unknowns and I think they did a really good job of highlighting those and making some assumptions about the future. Again, we don't know for sure, but these are, at least based on fact and their own research versus just theory.
Paul Roetzer: And I think it is a, it's a topic we're probably going to come back to a lot next year as things start to play out. But the questions you start to wonder is when there is just a generative AI output in a search engine and it, maybe it has some links or citations, where are those coming from? If not, my personal experience has been they are not the same [00:55:00] quality as the top 10 results on a normal search engine result page.
Paul Roetzer: I've had instances where it gives me, links and I click through them and just junk, scam, spam sites. And these are coming from Google and Microsoft. So something's off, the algorithm with how it would, how selecting what to inject into those results is weird.
Paul Roetzer: It's not like their normal algorithms, which I assume they'll solve, but that's an unknown right now. It's how? Where are those coming from and how do you get in those links if doing what we did previously to get in the top three to five links on page one doesn't get us in these generative AI responses, then what do you do differently?
Paul Roetzer: I think watching very closely organic traffic. And then just overall, what, what does that mean to our strategy? how do we evolve our content strategies if all of this ends up being true, if we see a big drop in organic traffic, if we see a change in consumer behavior, I've said on this show before, clicking on the links on [00:56:00] Google feels obsolete.
Paul Roetzer: going to a search result page and having to work for the answer feels weird to me now. and when that happens, that's when, change is coming. That's when consumer behavior is altering when you do an experience that feels archaic. And so I just, I think it's inevitable that we, the search looks different next year, it's going, it's going to be a different experience.
Paul Roetzer: We just don't know how it plays out. So I think all marketers, all business leaders really need to be paying attention to that, especially if you rely heavily on organic traffic and add revenue from that traffic.
Mike Kaput: last but not least on our docket today, Google just confirmed that it's investing up to 2 billion in AI company Anthropic, the maker of the Claude chatbot.
Mike Kaput: Google has already invested 500 million into the company and is upping that by 1. 5 billion over time. Now, this comes right on the back of Amazon's announcement last month. [00:57:00] That it would also invest 4 billion up to 4 billion into Anthropic Paul. what, what does Anthropic doing to deserve this windfall?
Mike Kaput: It's been a good couple of months.
Paul Roetzer: Yeah. A major player in the language model space. , big focuses on safety. And so they claim, Dario Amodei, we did a podcast episode. We talked about one of his interviews extensively, and he came from, safety and security at OpenAI to found.
Paul Roetzer: And anthropic, their belief is they have to build extremely powerful models to figure out how to protect humanity from them. And so racing to build powerful models to figure out how to put guardrails around them. it's sort of the catch 22 right now is the companies that want to build safety and security into these have to build the models first and then building the models there's risk.
Paul Roetzer: So , just, in a very important company. we'd say it often, you have, you have Google, you have. Microsoft, [00:58:00] you have Anthropic, you have Inflection, OpenAI, these are the companies that will define the future of humanity in many ways, especially if The government regulations we talked about earlier, protect the leaders.
Paul Roetzer: They become even more powerful and important to the future of humanity and certainly of business at a smaller level. So yeah, it's a company to pay attention to.
Mike Kaput: All right. Well, believe it or not. There were almost a dozen topics that didn't even make it into this podcast episode, and this is why we have revamped the format of the Marketing AI Institute's newsletter, which goes out weekly.
Mike Kaput: If you go to marketingaiinstitute.com/newsletter, you can join more than 25,000
Mike Kaput: next gen marketers who get our newsletter every week. It goes out on Tuesdays, and we've revamped it to include To make sure it is this week in AI. So we are including all of the news [00:59:00] and analysis we discussed both on the podcast and the stuff that doesn't make it onto the episode. So you can in one weekly digest, get a full picture of what's going on in AI every week, which I'm sure many of our listeners know, isn't always easy to do on your own.
Mike Kaput: So I'd highly encourage you to check out marketingaiinstitute. com forward slash newsletter to check out. that weekly digest and Paul, as always, thank you for breaking down what's going on this week in AI. We really appreciate it.
Paul Roetzer: Yeah, it's, it's good stuff. And we were thankful for everyone listening.
Paul Roetzer: And again, encourage you form your own perspectives on this stuff. We try and give us balanced, of viewpoints as possible on all these topics. Hopefully we do a good job of that. when there is bias, we try and illuminate what our bias may be on these topics. But our whole point is to try and give you the knowledge and resources so you can form your own opinions and hopefully help [01:00:00] ensure that we have the best possible outcome from AI.
Paul Roetzer: And we, we have a human centered approach to it and everything we do. So thanks for listening. We will talk to you again next week. Thanks for listening to the Marketing AI Show. If you like what you heard, you can subscribe on your favorite podcast app, and if you're ready to continue your learning, head over to www.marketingaiinstitute.com. Be sure to subscribe to our weekly newsletter, check out our free monthly webinars, and explore dozens of online courses and professional certifications.
Paul Roetzer: Until next time, stay curious and explore AI.
Claire Prudhomme is the Marketing Manager of Media and Content at the Marketing AI Institute. With a background in content marketing, video production and a deep interest in AI public policy, Claire brings a broad skill set to her role. Claire combines her skills, passion for storytelling, and dedication to lifelong learning to drive the Marketing AI Institute's mission forward.