Over the past decade, compensation for artificial intelligence (AI) professionals has surged at an unprecedented pace, reshaping the talent market and redefining what employers must offer to attract and retain top-tier technical talent. As companies across nearly every sector race to integrate machine learning, automation, and generative AI into their operations, the demand for skilled AI engineers, researchers, and product leaders has vastly outstripped supply. The result is a compensation environment that is not only highly competitive, but increasingly aggressive.
What makes this shift especially striking is how rapidly it has accelerated. Even five years ago, AI roles commanded above-average compensation, but nowhere near the levels seen today. Now, seven-figure packages for senior AI experts are not only possible, theyβre becoming increasingly common.
This surge is driven by a unique convergence of market forces: the explosion of generative AI capabilities, a shortage of qualified talent, escalating corporate reliance on AI strategy, and the emergence of new startup and investment ecosystems flush with capital. Together, these factors are pushing AI compensation to historic highs, with no signs of slowing down.
And of course, this article was written with the research assistance of AI.
AI is one of the few fields in which global demand massively exceeds global supply of qualified professionals. Only a small subset of software engineers possess the deep expertise required for advanced machine learning, reinforcement learning, natural language processing, and large-scale model development. Even fewer have hands-on experience with cutting-edge deep learning architectures or the ability to integrate foundation models into commercial products.
Companies are discovering that they are effectively competing for the same limited pool of elite talent. And that competition is fierce.
Here are a few key reasons AI talent is scarce:
This scarcity alone would elevate compensation, but the explosive commercial potential of AI has supercharged it.
The release of large-scale generative AI models has catalyzed a gold rush. Companies of all sizes now recognize that AI will determine competitive advantage in the coming decade. As firms shift from βAI experimentsβ to βAI strategy,β the urgency to hire expert talent has become acute.
Generative AI has created entirely new job categories, including:
In many cases, these roles did not exist 18 months ago. Now, they are some of the highest-paying jobs in the technology sector.
Compensation varies widely based on geography, seniority, company size, and specialization. But one trend is clear: AI salaries are increasing across the board, often dramatically.
Typical U.S. salary ranges for AI roles:
And these figures do not account for extreme outliersβmost notably the seven-figure offers made by OpenAI, Anthropic, Google DeepMind, Meta, and specialized hedge funds or trading firms.
Compensation for AI talent is highest in the Silicon Valley/San Francisco area, followed by New York and then Seattle.
AI startup funding is booming. Investors are pouring billions into companies developing foundation models, AI infrastructure, and vertical AI applications. With capital plentiful and competition intense, startups are offering generous equity to lure experienced AI hires away from Big Tech.
What startups are offering:
For highly specialized engineers, particularly those with LLM or multimodal model experience, equity stakes can be extremely significant.
The big players are stepping up as well. In late 2025, OpenAIβs average stock compensation reportedly reached $1.5 million per employee for its 4000 person workforce.
AI is no longer limited to technology firms. Industries such as healthcare, finance, manufacturing, retail, defense, and media all have aggressive AI build-out strategies. This has expanded the competition for talent beyond Silicon Valley, creating upward pressure on compensation.
For example:
These companies often have substantial cash reserves, enabling them to offer compelling salary packages more commonly associated with Big Tech.
Remote-first hiring has created a global bidding environment. Companies that once paid lower regional salaries are now forced to match global standardsβespecially when competing against deep-pocketed AI enterprises and venture-backed startups.
As a result:
This globalization has further driven compensation upward.
As poaching becomes rampant, companies are creating elaborate retention structures, including:
Companies recognize that replacing a senior AI engineer or researcher is extremely costly, and often impossible in the short term.
Companies should expect:
Organizations that fail to invest in AI talent will struggle to compete strategically, technologically, and operationally.
For employees, the moment is historic. AI expertise, especially in LLMs, applied machine learning, infrastructure, safety, and AI product design, is one of the most valuable skill sets in the global economy.
Professionals should:
Those with the right skills can expect strong compensation growth for the foreseeable future.
This section outlines the most important strategies, components, and negotiation techniques AI employees can use to maximize compensation and secure long-term professional protection.
A common mistake candidates make is focusing on base salary alone. In AI rolesβespecially at high-growth startupsβbase salary may not be the most important part of the package.
AI employees should evaluate:
Total compensation packages in AI can vary by hundreds of thousands of dollars depending on equity and incentives, making it essential to evaluate the full structure.
AI startups and AI-first public companies rely heavily on equity to attract top-tier talent. But equity terms are nuanced and highly negotiable.
Key equity terms you should negotiate:
A single percentage point of equity at a strong AI startup can be worth millions of dollars in a successful exit. Do not underestimate your ability to negotiate this component.
Pro tip: Ask for your equity in terms of percentage ownership, not number of shares. This forces companies to reveal the fully diluted share count.
AI work is often tied to quantifiable outcomes: model accuracy, latency improvements, deployment milestones, or product releases. This makes it easier to negotiate objective bonus structures, rather than subjective or discretionary ones.
You can negotiate:
Beyond salary and bonuses, benefits protect well-being and support work-life integrationβparticularly important for senior leaders.
Benefits can include:
Given the velocity of change in AIβfunding cycles, pivots, acquisitions, and leadership turnover, severance protections are essential. They are highly negotiable for AI professionals.
Negotiate for:
Many AI companies do not offer severance by default, but will add it if asked by a senior or highly valuable hire.
AI employees who interview with multiple companies often have dramatically better outcomes. Even one additional offer can significantly increase your negotiation leverage.
Tips for handling competing offers:
Companies expect AI talent to be in high demand. You should expect and encourage competition.
AI work often includes high-stakes systems, regulatory exposure, or sensitive data. Professionals should negotiate strong protections.
You can ask for:
AI professionals involved in model development, compliance, or safety can insist on explicit liability protection.
AI talent is global, and many companies are remote-first. If location flexibility matters to you, negotiate it early.
You can request:
Given how scarce AI talent is, many companies will accommodate flexibility for the right candidate.
Here are some additional important issues to consider when negotiating an employment contract or offer letter:
AI compensation packages, especially those involving equity, are increasingly complex. Understanding tax implications, vesting schedules, and contract terms often requires professional review.
An attorney or advisor can help you:
A modest legal investment can protect hundreds of thousandsβand sometimes millionsβof dollars in future compensation. And sometimes you can negotiate for the company to reimburse your reasonable legal fees incurred.
AI employees today are in a uniquely powerful negotiating position. Compensation is skyrocketing. Companies are racing to hire scarce talent, and the strategic importance of AI expertise has never been higher. By approaching negotiations with clarity, confidence, and a deep understanding of total compensation, AI professionals can secure packages that reflect both their current value and their long-term contribution.
In an era defined by rapid innovation and intense competition, negotiating well is not just a financial decision, itβs a strategic career move.
Related Articles:
Glass Imaging's impressive GlassAI Neural image signal processing (ISP) technology is heavily featured in the brand-new Honor 600 smartphone, promising to improve the zoom photography experience on the fancy new phone.
The central government has greenlighted yachts from Hong Kong and Macau to enter nine mainland Chinese cities in the Greater Bay Area (GBA) without customs guarantees and with temporary ship registrations.
The State Council said on Friday that the pilot scheme came into effect the same day.
The nine cities are Guangzhou, Shenzhen, Zhuhai, Foshan, Huizhou, Dongguan, Zhongshan, Jiangmen, and Zhaoqing.
The State Council also said it had βtemporarily adjustedβ two maritime ordinances to relax restrictions for Hong Kong and Macau yachts travelling to the Greater Bay Area.
The Hong Kong government said on Saturday that it welcomed the new policy.
βUnder the new policy, the exemption [from] the requirement for a guarantee will significantly reduce the financial burdenβ on yacht owners, it said in a statement.
It also praised the simplified registration scheme, which allows Hong Kong and Macau yachts to obtain temporary national ship registration from mainland China βwithout affecting their original ship registration.β
Owners of Hong Kong and Macau yachts previously had to pay hefty customs guarantees and undergo complex registration procedures before entering mainland ports.
According to the Saturday statement, the Hong Kong Marine Department will βcontinuously review and refine the facilitation measures for the northbound travel of yachts from Hong Kong and southbound travel for yachts from the Chinese Mainland to foster a healthy, sustainable and competitive environment for the development of the local yacht economy.β
Hong Kong Chief Executive John Lee first proposed boosting the local yacht economy in his 2025 Policy Address in September.
The following month, the government announced new initiatives, including providing 600 additional yacht berths at the former Lamma Quarry, expanding the Aberdeen Typhoon Shelter and the Hung Hom Station waterfront projects, and a planned yacht bay at Airport City.
FolsomNatural posted a photo:
Black & White public domain photographs colorized using ArtSpace AI software, with additional photo processing.
Coastal landscapes are constantly being reshaped by natural forces, and as climate change causes more frequent storms and sea level rise, that change will only intensify. Because these areas are densely populated with homes, tourist destinations, and industries, understanding how and where the coast will change is a pressing issue. However, reliable predictions that lead to actionable knowledge are rare.
Lentz et al. describe the state of knowledge regarding coastal evolution, highlight gaps in scientistsβ understanding, and describe opportunities for integrating information from various models, data sources, and end users.
Current coastal evolution predictions are often focused on too specific a location and are therefore hard to generalize or analyze too large a region and therefore lack detail, the authors say. In addition, itβs challenging for researchers to link the effects of acute events, such as storms, with long-term trends like sea level rise.
Improving these simulations will likely require combining many different types of models, including physics-based numerical models, models based on empirical measurements, and statistical models that include machine learning. To fully understand potential changes, the authors note that it is also essential to consider both coastal processes and human actions.
The researchers recommend several ways to improve consistency and collaboration in the field of coastal change forecasting. First, standardizing approaches and outcomes would make it easier to produce national-scale predictions. Right now, the variety of tools used across different locations makes it difficult for scientists to compare results and communicate effectively. They also emphasize the need for using coordinated research approaches. Stronger transdisciplinary collaboration, accompanied by essential training and support, would also enable scientists to make better predictions, the researchers say.
Comparing predictions to real-world observations of coastal landscape change could also help untangle this multifaceted challenge. By studying how coastlines have already changed, researchers can validate models and choose those that are performing best. Such comparisons require datasets that adequately capture coastal landscape change across both time and space. Remote sensing data and the use of artificial intelligence (AI) for data processing may help provide these improved datasets, the researchers suggest.
Engaging end users during the project planning process is also helpful because only end users truly know what kind of information they need to adapt to landscape change. Knowing how to engage end users can be difficult for physical scientists, but various tools and specialized personnel exist who can help coordinate these interactions, the authors say. (Earthβs Future, https://doi.org/10.1029/2024EF005833, 2026)
βSaima May Sidik (@saimamay.bsky.social), Science Writer
OpenAI has announced that images generated with ChatGPT, Codex, and its API will include C2PA metadata and a SynthID watermark -- the two leading protocols in identifying AI images.
And yes, there is a reason Mercy is dressed that way.
With Japan being home to many of the worldβs most passionate gamers, Japan Railways Group periodically partners with popular video game franchises to add a little extra fun to fansβ train travel in Japan, like weβve seen in previous tie-ups with the Final Fantasy and PokΓ©mon series. For its next video game collaboration, though, JR Central (a.k.a. JR Tokai) is teaming up with a series that hails from the U.S.
This summer, characters from the cast of Blizzardβs Overwatch will be taking a ride on the Shinkansen as part of the Payload to Osaka campaign (a reference to the gameβs βpayload escortβ missions). The promotionβs key art shows Hanzo, D.Va, Genji, Kiriko, Mercy, Hammond, and Jetpack Cat getting off the bullet train at Shin Osaka Station, the closest stop on Japanβs high-speed rail network to downtown Osaka. The artwork contains a few tips of the hat to Osakaβs famous food culture, with Hammond snacking on takoyaki octopus balls and Hanzo scarfing on a steamed pork bun (and youβll note that Hanzo is eating it after getting off the train).
From July 17 to September 23, passengers onboard JR Central Shinkansen trains can take an Overwatch-themed quiz via their smartphones, and correctly answering the questions gets you one of seven phone wallpapers. While the wallpaper you win is randomized, you can take the quiz as many times as you want during your ride (10 questions are randomly pulled from a pool each time), and JR Central guarantees you wonβt get any duplicate wallpapers until youβve completed a full set of all seven.
Even if youβre not up to taking the quiz, you can still receive a PC wallpaper image of the illustration just by answering an online questionnaire, which, like the quiz, can be accessed through the promotionβs official website, while onboard the bullet train. This also serves as a record of your Shinkansen ride, which you can then show at the Osaka Nipponbashi branch of anime merchandise store chain Animate to receive an Overwatch cleaning cloth with the same illustration.
Animate Osaka Nipponbashi will also be hosting an Overwatch Payload to Osaka popup store with character pins and acrylic mini standees from July 17 to 26.
While Overwatch does have fans in Japan, the seriesβ most passionate supporters tend to hail from other countries. With international tourists accounting for so much travel in Japan these days, though, especially on the JR Tokai section of the Shinkansen network that connects Osaka with Kyoto and Tokyo, JR Central most likely expects to get a lot of interest from non-Japanese fans, enough so that the company felt the need to follow up on its official Twitter announcement for the collaboration with a statement that βThese items are available in Japan only.β
Speaking of the Twitter announcement, at least one non-Japanese fan was furious about the clothing that Mercy is wearing in the artwork.
What is this kind of OUTFIT? ugly as hell pic.twitter.com/4wkggD4fpR
β π©β‘πͺ
From the choice of words and capitalization, that was probably meant as a rhetorical question, but hey, thereβs actually a reason for this outfit (sorry, OUTFIT). Aside from having a vibrant food culture, Osaka is also known for its flashy fashion sense, in particular its ladiesβ fondness for animal prints. Granted, itβs usually associated with women of a more mature age than Mercy, but a little touch of Osaka style isnβt completely out of place here.
Related: JR x Overwatch Payload to Osaka official website, Animate Osaka Nipponbashi
Source: JR Central via Hachima Kiko
Images: JR Central
β Want to hear about SoraNews24βs latest articles as soon as theyβre published? Follow us on Facebook and Twitter!
The Role of AI in Modern Copywriting written by John Jantsch read more at Duct Tape Marketing
OverviewIn this episode of the Duct Tape Marketing Podcast, John Jantsch sits down with Jon Benson, creator of the Video Sales Letter (VSL) and founder of the AI platform Benson. Jon shares how AI is reshaping the world of copywriting, not by replacing human creativity, but by amplifying it.
The conversation explores the evolution of VSLs, why they continue to outperform despite industry skepticism, and how AI is changing the way marketers create, test, and optimize content at scale. Jon also dives into the importance of maintaining a human voice, building ethical persuasion frameworks, and avoiding the trap of generic AI-generated content.
Jon Benson is a copywriter, entrepreneur, and AI innovator best known for creating the Video Sales Letter (VSL), a format that revolutionized digital marketing. With a background in persuasion and behavioral psychology, Jon has spent decades refining ethical copywriting techniques. He is the founder of Benson, an AI platform trained on high-converting campaigns designed to help businesses create more effective, human-centered marketing.
The real opportunity with AI is turning marketers into better editors, strategists, and decision-makers, not eliminating the human role.
Despite claims that theyβre outdated, VSLs continue to drive strong results when built on solid messaging and persuasive structure.
Whether itβs video, text, or ads, the effectiveness of marketing still comes down to the quality of the words and messaging.
Generic prompts produce generic results. AI needs context, personality, and values to generate effective copy.
Great marketing aligns with what customers already believe and value, rather than trying to force persuasion.
Top marketers run hundreds of variations simultaneously, something only possible at scale with AI.
Without clear boundaries, AI can drift into manipulative messaging. Defining what to say and what not to say is critical.
Like upgrading from a hammer to a power tool, AI removes manual effort so humans can focus on higher-level creativity.
To get quality output, users must teach AI their voice, values, and audience rather than relying on default behavior.
Even with AI, understanding persuasion fundamentals and customer psychology remains essential.
00:01 β AI as a Creative Multiplier
John introduces the idea that AI enhances, not replaces, human creativity.
01:16 β The Birth of the VSL
Jon shares how Video Sales Letters transformed his career and the marketing landscape.
04:08 β Early Adoption of AI in Copywriting
Jon explains his long-term vision for AI-powered copy tools.
06:21 β Are VSLs Overused?
Why VSLs continue to perform despite years of skepticism.
08:46 β Why Words Still Win
The importance of messaging over format in marketing success.
09:11 β The Problem with Generic AI Content
Why most AI-generated content feels robotic and ineffective.
11:40 β The Role of Personality in Copy
How values and voice shape better marketing outcomes.
14:26 β AI as a Creative Partner
Using AI to enhance, not replace, human creativity.
16:37 β The Power of Testing at Scale
How AI enables massive experimentation and optimization.
18:23 β Ethical Guardrails in AI Marketing
Why defining boundaries is essential for responsible persuasion.
βThe words are the consistent thing. If the words donβt reflect a human, people sense it immediately.β
βAI isnβt the answer, itβs a tool. You still need to bring strategy and voice to it.β
βYouβre not trying to convince people, youβre aligning with what they already value.β
βThink of AI as a power tool, it removes the grunt work so you can focus on creativity.β
John Jantsch (00:01.651)
So what if the real opportunity with AI is not replacing human creativity but expanding it by turning entrepreneurs into better editors, directors, and decision makers? Hello and welcome to another episode of the Duck Tape Marketing Podcast. This is John Jantsch. My guest today is Jon Benson. He's a copywriter, entrepreneur, and AI pioneer best known for creating the video sales letter, one of those terms that people just use like it's been around forever.
A format that shapes modern digital marketing. is long centered on ethical persuasion and authentic connection. And more recently, he developed BNSN, an AI platform trained on high converting campaigns for small businesses. So John, welcome to the show.
Jon Benson (00:29.9)
Yeah.
Jon Benson (00:47.212)
Hey, John. Thanks for having me.
John Jantsch (00:49.585)
So let's, I assume you have to do this a little bit of your time when you go on shows like this, but the term VSL, you know, is kind of entered the, the marketing vernacular. Talk to me a little bit about, I've been doing this for 30 years. That was probably 12, 15 years ago, really, when that kind of burst on the scene as an innovation. You want to talk a little bit about what that's done to your trajectory, I suppose.
Jon Benson (00:55.202)
Mm-hmm. Yeah. yeah.
Jon Benson (01:02.04)
Mm-hmm.
Jon Benson (01:16.216)
Yeah, believe it or not, it's 20 years old this year. So 2006. Yeah. Yeah. Crazy. It's, mean, it was, it, yeah, everything changed that the, day that happened, the 30 days later, everything changed from my offer that I did it for, you know, we went from like struggling onto my second book that I wrote in, in fitness and then went to a million dollars.
John Jantsch (01:18.537)
20 years, okay.
Jon Benson (01:39.886)
a week and a month rather in traffic cost, you people buying that kind of money and going up to even higher than that. So it was crazy. And then, and then all of people started calling me and asking me to write VSLs for them. And I'm not, I wasn't a copywriter. that's not, never been my claim to fame until after this happened. And then I had to get good at writing copy. So that's what happened.
John Jantsch (02:01.939)
That's funny. So you said you had written a book about gym ownership? Is that what you said?
Jon Benson (02:10.663)
I've written six books in fitness, so weight loss, fitness, bodybuilding, yeah, so that whole thing has been a passion.
John Jantsch (02:12.947)
Fitness, fitness, okay. Okay, so are you one of those people that that was your passion and you just had to learn how to do marketing? And so this whole idea of studying persuasion and conversion and innovation, is that something that was really just picked up because you're like, I better get good at that?
Jon Benson (02:24.748)
Yeah.
Jon Benson (02:34.478)
It was picked up specifically for copywriting, yes, but I studied persuasion in college. Actually, I was studying MLP in college. I was fascinated by how you can basically get people to listen to you and hear what you're actually trying to communicate and motivate them to make changes based on things that you believe at least are good for them. So you're not trying to manipulate them. You're just trying to motivate them. And I was always into like, how can I motivate and connect with people deeper? So I studied the MLP back then, way back then.
John Jantsch (02:39.731)
Mm.
Jon Benson (03:03.22)
and mail order course from, from Bandler. And that got me into Tony Robbins and that led me into even deeper persuasion issues. And, and just was always really fascinated by it. And that led to me being into the advertising world. And that would, that led eventually to writing a book with it. Yeah. I actually would have the book thing came about because I'd always been passionate about, bodybuilding and fitness and things like that growing up and athlete. I was an athlete most of my life. And then
ended up sedentary and got ended up obese in my late 20s and early 30s. I had 50 inch waist and had a heart attack at 38. So I was like, it was like a train wreck of health. And that got me back into it. So that's the Fit Over 40 book was written based on that, on turning that around. And then I interviewed a bunch of other people because I didn't think I was enough for a book. So I did 52 people that did the same.
John Jantsch (03:55.283)
So I'm curious, this is a question, unfortunately, I feel like I'm asking almost every guest these days, but how has AI changed that element of copywriting for good or bad?
Jon Benson (04:00.942)
It's
Jon Benson (04:08.494)
So my goal with AI and copywriting, I've been doing copywriting software since 2010. So this is going to date me a lot, but in AI, in early nascent AI in 2017 and working with early LLMs in 2019. So very, very, very early into this thing and trying to convince everybody, this was the thing that we wanted to do. And the reason why is because I was, I had these courses that I would teach people how to write VSOs and I knew how hard it was for me to learn all the copywriting in and outs and
and develop my own style, which I did. And I said, well, what, what if I could have software that would do it for them? And the average business owner doesn't have time to do that. They just want the copy that converts. So I've seen it from 15 years away going, I know this is going to happen eventually. And so we decided that the software is pronounced Benson. That's not my last name. It's just my last name without the vowels. And, and yeah, yeah, but it's, cool that you can spell it out. That's all right. and so we did Benson originally, it was going to be called,
John Jantsch (04:56.529)
okay. Not BNSM like I butchered it, okay?
Jon Benson (05:06.35)
It was going to, because it was the first AI to actually write a long form VSL. And I was working with, with Jasper at the time they were called Jarvis, but I was the first guy in the copywriter to train anything on an LLM. And they ended up with a 62nd VSL out of all the training. I think, yeah, I think we can do this in a different way. And we ended up being, you know, having a 7,000 word VSL come out of our AI and it sounded like a real VSL.
John Jantsch (05:14.729)
Sure, yeah.
Jon Benson (05:32.663)
It didn't sound like chat, GBT, it didn't sound like Claude, it sounded like a real VSL. And so that was our claim to fame. And since then we just, of course got, we were very early into the agentic phase. So we've just gotten better and better at that. And so my goal was to replace myself. That's what I wanted to do. I wanted to say, if I can, if I can use this to write a VSL, which I have, sells pages for my own stuff, which I have, then I know that it's going to be good enough to, for prime time. And that was the, that was the goal to do. yeah.
John Jantsch (06:02.549)
So talk to, obviously we've got more to explore in AI, but talk to me a little bit about the VSL itself. mean, it has become very mainstream. I mean, you hear people talk about it, whether they know what it is or not. They talk about it as part of their funnel, you know, today. So is it overdone? I mean, is it over?
Jon Benson (06:06.094)
Mm-hmm.
Jon Benson (06:10.316)
Mm. Yep.
yeah.
Jon Benson (06:21.806)
Yeah, every year I hear that I've heard that for 20 years. So it literally 20 years. So the first year I came out with it and said, Oh, it's already and then Ryan Dias, who's a good friend of mine made the mistake of saying when he came out and promoted his own little mini VSO course and he later gave me credit for which was really nice of him and everything. But he said, Oh, sales letters are dead. You'll never do another sales. And I'm like, dude, I've never said that, you know, I think everything works if you let it and VSO is just happened to keep on working and they just ask, ask Agora.
John Jantsch (06:24.157)
Yeah. Yeah. Yeah.
Jon Benson (06:51.022)
They work. I mean, yeah, they work. They work really well and now people are using BSLs in feed So you've got the meta ads that are basically short BSLs that use the same psychology Just compressed into five two to five minutes. So we've been doing that for 15 years as well So yeah, and then they go to a longer BSL So they they still work just as sales pages work just as webinars can work everything can work It just depends on what you're wanting to sell and how you're and how you approach it But the words are the consistent thing
So if the words aren't there, if the words don't reflect an actual human underneath it, people sense it a mile away, which was our goal with Benson was to create humanized AI. How do we do this? How do we create AI that doesn't sound robotic? It doesn't sound like, you know, chat GPT writing an email, it's asking a rhetorical question. And the very first sentence, you know, this kind of really bad AI copy that we see all the time. How do we do this and actually sound like a real A-list copywriter? And that was, that's been our focus for three and a half years now.
John Jantsch (07:20.456)
Yeah.
John Jantsch (07:48.413)
You know, initially the large innovation was that it was not a talking head on video. It was the words. Is that a key component of it?
Jon Benson (07:56.174)
Mm-hmm.
Jon Benson (08:01.113)
You know, it depends on what you're trying to sell. We have seen split tests with video beating words only, and we've seen words only beat video. It really depends on what it is. And what works today, a year from now, will be something you want to reverse. So for a while there was like my friend Craig who writes for Golden Hippo, and he's done amazingly well building a billion dollar company from, he's an amazing writer. But he was one of the first guys working with Gundry to do a lot of video.
on the front end of a VSL, but talking to him behind the scenes, so to say, we know that it's still like a Google Doc and the words are everything. So he slaves over the words, man, getting the words just right. So all the video in the world is not gonna save you if your words suck. It just isn't gonna happen. So the words are still the most important.
John Jantsch (08:46.077)
Yeah, yeah, yeah.
So one of the knocks on AI, of course, is it's made it very easy for people to create really crappy content. you see it all the time now, right? It's like volumes of really bad content. So why can't people create better content? What's the mistake they're making? Is it simply just a matter of being lazy?
Jon Benson (08:54.831)
Yeah, yeah.
Yeah.
Yeah.
Jon Benson (09:11.983)
No, it's the matter of the LLMs or the in our case, it's the agents not knowing you. And this is where it gets a little bit a little bit hairy for people, because there has to be an element of your personality that's OK to be known. as the same thing would be true if you went and hired me as a copywriter. Like I would ask you if you had an offer and you wanted to whatever your offer would be. I would start asking you lots of questions that you probably don't think is related to your offer.
John Jantsch (09:19.719)
Yeah, yeah,
Jon Benson (09:40.336)
Now I'm not talking about like when asking all these really intensive personal questions, but I want to know what your values are. I want to know where you stand. Who do you want to attract as customers? What are you against? What are you not just what the, what the product does? Cause the product or the offer, whatever it does, I that's, that's not that difficult. Um, what's difficult is to make that story resonate with people that will automatically hear and go, Oh, that sounds like something that I can automatically relate to. And that's what a really good copy. does. We don't try to sell people that are
not interested or just completely need to go from a level one to a level five awareness, that's really not what we wanna do. We wanna target people that are already there, because you got plenty of people like that, but if you write, if you go into a chat or clod or whatever and you say, write me an email or write me an ad or rep me a VSO, and they don't know who you are, they don't have a good feel of your words, feel of your personality, it's gonna write stuff that's schlocky, because it's trained on the internet. So if you just think about this for a moment, and everyone listening to me will get this,
John Jantsch (10:35.294)
Yeah, yeah.
Jon Benson (10:39.043)
It's like, can you imagine training anyone to do anything by telling them, go read the internet and get back to me tomorrow? That's what we've done with LLMs, right? It's like, well, that's going to give you a lot of knowledge, but most of it sucks. mean, so most of what's out there in copy is terrible. So it's learning models have been terrible. So that's why specialty AI is like ours and in our, in our industry, you have to have it to where the people that know what they're doing actually trained individual.
John Jantsch (10:46.665)
Right.
Jon Benson (11:06.487)
in our cases, agents that use not one LLM, but a dozen, you know, can use as many as we need one model rather, but you know, doesn't whatever models are we know are going to be the best ones for the right tasks. So that takes that. And then what we do is a little different. We ask people to go through an assessment to figure out what are their values? Where do they stand? Who are the people they want to attract? And how do they want their their words to appear? So we take care of the persuasion element, but also we see that with the words and phrases that
John Jantsch (11:14.739)
Yeah, yeah.
John Jantsch (11:25.885)
Mm.
Jon Benson (11:35.681)
are closer to who they are as a person. So it starts feeling more human. It's important.
John Jantsch (11:40.457)
Yeah, it's interesting. know as we've worked with clients, you know, a lot of them have a fairly large body of work of them talking about things, explaining their products, being who they are. And that element, you know, allows you to build that voice or that brand. But then there is a technical framework element to it as well, isn't it?
Jon Benson (11:58.348)
yeah, totally. mean, if you go too far outside that framework, you're going to lose a lot of the things that we already know work so well, persuasion wise. So the goal is not to try to convince somebody of something, it's to compel them to take action on what they already hold valuable. So all you're doing is aligning your offer with what they already hold to be valuable. And that's the skill of copywriting. that's something that AI is, I think, obviously I'm biased.
John Jantsch (12:05.639)
Yeah. Yeah.
Jon Benson (12:27.481)
So I'm gonna say we're kind of the exception, but AI in general has gotten a little better at this. I'd like to think we've led some of the way in that, to getting to where there's more of that human element involved.
John Jantsch (12:39.091)
So talk a little bit about that because there's certainly a lot of people, creatives in particular, that have felt like they have this special sauce, this special talent to create that content, to create beauty, to create things. And maybe AI has kind of taken that. I mean, it's eventually going to get good at doing video and graphics and things. So where is the human element, know, remain?
Jon Benson (12:57.314)
Mm-hmm.
Yeah, yeah.
So think of it as like, I look at it as the difference between using a hammer and using a jackhammer or something that's a powered hammer, right? It's a pneumonic hammer or whatever they call those automatic hammers. So you've got an automatic hammer and there's a skill to hitting a nail with a hammer, right? The question is, as a carpenter, is that really what you want to be known for is I strike a nail head perfectly with a hammer every single time.
Or if you could have that done for you instantaneously with something that just tapped it in, what would you do with the time that you have left now? You would probably spend that doing the creative portion of things and like, I can do this, I can build this. And this is what the same thing is true of AI and copywriters. It's like, we're not trying to put people out of business. We're giving them the ultimate power tools. So a lot of the grunt work, a lot of the research, a lot of the structure you don't have to worry about. Then you can go in and finesse it.
and everything sounds so much better when you do that. We want people to do that. there's still a knowledge factor that I think that copywriters need to have. And sure, some people do use tools like Vinson. They just don't think about it. They click a few buttons and they go, because it works. But the copywriters, they want to put their signature on it. And this just gives you the ultimate way of doing that. It's like hiring the best ghostwriter you can think of. So if I hired a copywriter to write something for me and they sent it back and I read it, went, wow, that's just freaking fantastic.
Jon Benson (14:26.768)
then I could find these little bitty things in there that I only know or that I primarily know. And then I'm gonna go, oh, you I'm gonna change this over here. And then I might find a creative thing that he said or she said that I wouldn't have thought of. And that now becomes a campaign. My mind goes, oh, wow, I didn't think about that. I can turn this into a campaign. Well, that's not AI, that's me, right? So if the AI wrote it or a human wrote it, wouldn't matter. And so that's what we do that's a little different because we coach people live once a week so that we can help inspire them to.
Use the words that are coming out and how can we use it to help market their business more effectively.
John Jantsch (15:01.011)
So I think one of the areas that obviously is a breakthrough is in testing. Obviously, any copywriter worth their salt is like, I think this is good, but let's test it, right? And now we can test 200 versions for not much more time than it took us to create that one beautiful one. What do you think that that is going to ultimately do in terms of people's effectiveness?
Jon Benson (15:07.088)
Mm-hmm.
Jon Benson (15:15.087)
Right.
Jon Benson (15:26.992)
If people knew what the guys that are making hundreds of millions of dollars at this stuff do, if you knew the amount of testing that went into it, most people would just give up. would stop. I'll give you an example. I have a good friend of mine that is the top of their industry on meta and they flew out to meet the actual real meta heads of ads because there's the ones that they give people and there were ones that give these people.
You know, they give them $100,000 to spend just to play with just because we want to see what your new creative team can do. They will run 800 ads at a time in any given month. They're running 800 versions of an ad. So there's just no way to do that effectively without AI. that's when they were the early adopters to this. Now they can run those kinds of things. And it's like, they can figure out what works and guess what? One or two might scale or three. It's, it's, doesn't matter how good the writers are.
It's like some hook, some angle may work and that angle if it works can just skyrocket a business. So I think it's one of the best things about AI is the ability to split test leads of a sales letter or VSL, the split test, obviously campaigns and then add campaigns and things like that. It's very helpful.
John Jantsch (16:37.907)
So you've spent a lot of time building a reputation about ethical persuasion, but it's not a very far leap to go to things that are maybe not that ethical, right? To go from just what you talked about as getting people to do something that they want to do or that's good for them and they just, they need to hear it, to manipulation. So, and I feel like
Jon Benson (16:43.12)
Mm-hmm.
Jon Benson (16:55.346)
yeah.
Jon Benson (17:01.796)
Right.
Right.
John Jantsch (17:07.503)
AI doesn't really care in some cases. how do you, what are the guard rails that you really use to kind of stay within what, you you talked about beliefs, your beliefs.
Jon Benson (17:10.072)
Mm-mm. Mm-hmm.
Jon Benson (17:20.24)
Yeah, well the guardrails I use that we actually that's a technical term and we use specific guardrails in our agents that are that when somebody sets up Benson correctly, we use it's called a buyer alignment profile that we have people go through. In fact, I'm going to give it to your listeners for free that could go through that and get their buyer alignment, which is a 15 page report of the words and phrases you should use and not use. And that exactly fits that bill of that sets up guardrails. It's like use this because I value X, Y and Z. What do the words of I
value X, Y, and Z translate to in copywriting lingo? Because it doesn't mean like if I value freedom, you don't want to use like, hey, since you love freedom as much as I do, then you're going to love so and so shoes. That doesn't make any sense, right? And so it's just too hamfisted and heavy handed and all that stuff. So what phrases do people that love freedom as a core value? What usage would they use and what would they never say? And it's what they would never say that the Garbrills of that. So in other words, that prevents the
John Jantsch (17:58.441)
All right.
Jon Benson (18:16.913)
AI from going over the balcony, so to say, when it comes down to overly persuasive language.
John Jantsch (18:23.251)
So for some of the folks that you've worked with, you've probably started to catalog kind some of the biggest mistakes people are doing, making right now using AI. Where do you see people really need to make a shift to make AI more effective for them?
Jon Benson (18:40.579)
it's it to stop thinking of AI as the answer and start thinking of it as a tool is a huge step in the right direction. Also to train whatever AI you're using. Ours is built to be trained, so it's copy paste kind of thing. But if you're going to use Claude or chat GPT or whatever, you need to be able to train it with who you are, what your values are, how what words or phrases to use, what not to use. And you'll find that the memory on this is pretty short. So.
unless you know what you're doing and then we can get into things like instances of open claw and the clawed code and all that stuff. That's very technical and most people don't want to go down that rabbit hole. mean, our guys go down that rabbit hole because we're kind of geeky when it comes to that. But most people want just the best answers that they can without having to become a software engineer. so to do that, yeah, it's a lot of knowledge. It's a lot of like time to say, here's who I am.
John Jantsch (19:08.713)
Mm.
John Jantsch (19:15.774)
Yeah.
John Jantsch (19:29.822)
me
Jon Benson (19:33.774)
And here's what I want you to do. Now, you can do that to a limited degree in chat and cloud and tools like that. You can do it to a huge degree in our tool because we built it to do that. And that's super important to get the language patterns down. But also, and this is the last thing I'll say, but this is true of copywriting in general. So when people used to hire me, because I don't write copy anymore. I'm solely focused on Benson. when people used to hire me, it was very expensive. I was like.
the probably the most expensive guy in the world for like five or 10 years. And they're certainly one of the most expensive guys in the world. And they would hire me and I would give them a first draft of something like usually a BSL or a sales letter. And they would say, this doesn't sound like me. go, yeah, I know. It's because you suck. Yeah, you don't want to sound like yourself, man. You really don't. it's and it's like, I, I mean, that in kind of a funny way. It's like you're the copy they were writing was just terrible.
And so they were trying to make their terrible copy kind of polish, you know, a poly put, put lipstick on a pig's episode. So you can't do that. You have to like be able to understand some basic persuasion and then work in. And this is what I didn't do when I was a pro when I was writing early days of copywriting work in their values. I figured this out later in my career. It's like, I can work in their value statements and figure out what the words are. But that was just tons of research. We'd charge like 15, 20 grand just to do the research to figure out like
John Jantsch (20:33.415)
Mm-hmm.
Jon Benson (20:58.491)
What are the words we should use and shouldn't use and phrases and all that stuff. And unless somebody came along that was like an identical client, we'd have to do that all the time. Now it's automatic, which is fantastic.
John Jantsch (21:06.473)
Yeah. Yeah. Yeah. Well, John, I appreciate you dropping by the duct tape marketing podcast. Is there someplace you mentioned that you had a gift you wanted to invite people? And obviously I'd love to know where they can find out more about Benson.
Jon Benson (21:15.471)
Yeah. Yeah. Sure. If you go to free buyer profile.com, that's free buyer profile.com. You can take our buyer alignment profile, which will test to figure out your core values, help you figure them out. We use a lot of different standardized testing models in these questions. And in about 10 to 15 minutes, we'll get you a report.
that you can use in your marketing that will tell you words and phrases that you should think about using and words and phrases you should definitely avoid. will give you all the NLP, all the magic sauce while still sounding like you and will also help elucidate what you already hold valuable and the people that
John Jantsch (21:53.481)
Great tool for training any AI tool, suspect, that you're going to use. Awesome. Well, again, I appreciate you dropping by. It's freebuyerprofile.com and hopefully we'll run into you one of these days out there on the road,
Jon Benson (21:57.125)
Yeah, definitely. Yeah.
Hmm.
Mm-hmm.
Thank you, John. I appreciate the time.
Enter your name and email address below and I'll send you periodic updates about the podcast.
Manuel Gual posted a photo:
Echoes of the Delta: A Visual Journey Through the Soul of the Blues
Description
This collection captures the raw, unyielding spirit of the American Deep South and the enduring legacy of blues music. From the mythical dirt crossroads under a midnight moon to the neon-lit doorways of bustling juke joints, these scenes weave a powerful narrative of hardship, resilience, and musical salvation. You can feel the steady rhythm in the weathered hands strumming a worn acoustic guitar on a wooden porch, hear the soulful wails echoing through smoky recording studios, and observe the quiet, melancholic contemplation in a lonely diner at dawn. It is a visual tribute to the pioneers of rhythm and blues, the rural landscapes where the sorrow songs were born, and the dusty highways that carried those melodies across the country. Every portraitβwhether it is a passionate singer pouring his heart into a microphone, lifelong friends sharing a laugh over a quiet evening, or a lone traveler leaning against a vintage pickup truckβembodies the authentic, vibrant pulse of Americana folklore.
These images were generated by Artificial Intelligence.
Login