• Bionic Marketing
  • Posts
  • Issue #17: Google’s AI hustle, Meta’s AI ads, 100K tokens, best AI creative suite

Issue #17: Google’s AI hustle, Meta’s AI ads, 100K tokens, best AI creative suite

Good morning.

It’s been weird watching Google struggle with AI lately. 

They have far more powerful AI systems behind closed doors than anyone else (tools that make ChatGPT look like a cute toy).

But between rumors of “code red”, all-hands-on-deck meetings, and an initial release of Bard that was kind of weak…

It looked like OpenAI and Microsoft were eating their lunch.

But the past week or so, Google has launched a flurry of AI apps and tools, named Duet AI, like:

  • The Google Search experience will soon come with AI answers prioritized at the top of the page. What will happen to SEM and SEO? That remains to be seen.

  • Launching PaLM 2 which excel at sense reasoning, mathematics, coding, and logic tasks. Some models can even run on your smartphone.

  • New photo-editing feature, Magic Editor. Change nearly every element of a photo, including adjusting lighting, removing unwanted foreground elements like backpack straps, and even moving the subject of the photo into other parts of the frame.

  • You can now turn text descriptions into music with MusicLM.

And, not wanting to be left behind, Meta makes Generative AI tools available to advertisers. 

They’re calling it the AI Sandbox and it’s like your own ads assistant, or media buyer, who can do anything you want to your ads.

You’ll continue to see tech giants and startups compete with integrating AI into everything.

Nothing will slow down. Everything will move faster.

But the trick is to not try to keep up with everything. There’s a lot of “me too!” features and tools right now. Most are copycats. Most of them don’t work that well.

Don’t focus on the tools or the news—be clear about what you need done, and find the AI tools that can do it for you.

In today’s issue:

  • 100K context window for Claude is a big deal.

  • Coca-Cola’s new AI ad (future of ads right here).

  • Warren Buffet weighs in on AI and the future.

  • Literally read the mind of your market with AI.

  • The best AI creative suite on the market right now?

Let’s dive in.


Anthropic, the company behind AI chatbot Claude, increased the context window with 100,000 tokens (via API):

This is far more than any other language model that’s available (though I expect that to change soon).

Why does this matter?

The context window determines how long your prompt can be (and the output you get back).

Right now, if you’re using ChatGPT, your context window is a few thousand tokens. Which translates to about 3,000 words or so. That’s apparently changing just the past few days, and OpenAI has expanded GPT-3.5’s context window by a few thousand tokens—which means you can fit more words in a single prompt.

To make it simple, a bigger context window means more use cases and capabilities.

You can do a lot more—and get more output back—without prompt stacking or chunking your input (which is what you mostly have to do right now).

Here’s what Anthropic said about their context window:

Put differently and in more human terms:

“The average person can read 100,000 tokens of text in ~5+ hours], and then they might need substantially longer to digest, remember, and analyze that information. Claude can now do this in less than a minute.”

If I were you, I would start using Claude for the context window (for now, via API—but soon available via their web interface) and try adding a whole sales page, or landing page, or several emails into one single prompt—and ask for whatever you want.

Ask it to rewrite a full sales page or landing page.

Ask it to analyze a sequence of 10 emails, highlight what’s good about them—and then ask it to rewrite them all according to the analysis.

Whatever marketing tasks you’re now doing—can be done with Claude in just a few prompts.

Or, really, with most AI tools out there.

It’s not just for text.

This is the future of AI and advertising. Either completely done by AI or a mix of human plus robots.

Is all this good or bad for marketers?

The conversation of good AI vs. bad AI rages on.

But the vibe shift in the market points to more acceptance of AI—and people turning their attention more toward how you can make money with it, or from it.

Mark Cuban highlighted his observation recently, comparing ChatGPT to the launch of the Internet:

And Warren Buffet gave his take on AI when asked about its disruption of the market at the 2023 Berkshire Hathaway annual meeting, saying:

“New things coming along don't take away the opportunities. What gives you opportunities is other people doing dumb things and I would say, in the 58 years running Berkshire, there has been a great increase in the number of people doing dumb things, and they do big dumb things and the reason they do it to some extent is because they can get money from people so much faster…

The big money… it isn’t outperforming and outperforming… if you're running small amounts of money, I think the opportunities will be greater.”

He even goes as far as comparing AI to the atomic bomb:

Even Sam Altman (CEO of OpenAI) recently shared some deep insights with MIT on the Imagination in Action channel where he discusses AI’s potential for humanity.

Those there were able to explore opportunities and dive into topics around AI-enabled solutions that solve real-time business problems.

The panel was bursting with pioneers, visionaries, and keynote speakers like Sam, who probed the future possibilities of AI on personalization, risk assessment, predictive analytics, and decision-making.

Speaking of MIT, they have their own mini-course for integrating AI into your business strategy. I’m sure we’ll see whole MBAs created on this over the next few years.

As always, I recommend that you just start using various tools to figure out what works. This is the best way to learn and prepare yourself for how much AI is already affecting work, business, music, arts—and everything you can imagine.


Imagine being able to create funnels, send emails, and even create a client logo with just your arm.

In a report shared by Science Daily, ACS Nano researchers devised a proof-of-concept silk armband that turns your forearm into a sketch pad or keyboard.

While this may not be true to what we typically associate with AI and marketing, this is the essence of Bionic: using machines to engage in marketing tasks and activities.

Don’t get stuck on what the intended use is. Imagine what else you can do with it.

The touch-responsive hydrogels are sandwiched in silk material so it doesn’t irritate the skin.

And like the keyboard or sketchpad, it interprets what a user creates or types and converts it to display on a computer.

What if your whole marketing campaign could be done with just a wave of your hand?

Speaking of body parts and AI:

As a marketer, Voice of Customer (VOC) research is essential to making sure whatever it is you are selling grabs the attention of those that need it.

Collecting data to connect with your audience, interviews, and message mining is necessary for making sure your messaging is what it needs to be.

If only we could read the minds of our market, right?

There’s an AI for that.

What do you get when you link a ChatGPT NLP with functional magnetic resonance imaging (fMRI)?

Real-time brain language decoders.

Now, these are not new. However, language decoders are typically invasive and require a type of neuro-surgery that limits their capacity to helping those who can’t move or speak, communicate.

After 15 years of research, human studies, and trials (and error), lead neuroscientist, Dr. Alexander Huth at the University of Texas combined the power of ChatGPT with brain recordings from fMRI to create the world's first “AI-based decoder that can translate brain activity into a continuous stream of text, [which] allows a person’s thoughts to be read non-invasively for the first time,” according to The Guardian.

Before, these recordings could only be used for limited processing due to a 10 second lag, and when used, were implanted into the brain.

Not anymore.

Not only have the combination of AI and fMRI scans brought forth a new way to record brain linguistics, they can now decode continuous natural language in real time, filling the 10-sec lag that has caused almost a two-decade gap in real-time language decoding—no implants required.

AI can now read your mind.

Pair that with MaMMUT, and we may be onto something even bigger:

The MaMUTT system has one vision part and one language part. It is able to learn different things at the same time in two steps with the language part.

This means it can use the same parts for different things, making it better and easier to use. It also makes it possible for this system to do object detection and video-language tasks more easily.

Are you thinking what I am thinking?

Brain to video.

Coming soon to a mind near you.

There’s money to be made in all things AI.


$100M: Runway. Since its start in 2018, they have raised nearly $200 million, with this most recent round taking it to that “nearly” mark. It’s no surprise as their user base consists of companies such as Google, VOX, NBC, Microsoft, and more. They are also the startup behind Stable Diffusion, and this last round of funding has boosted their valuation to $1.5 billion, ahead of its upcoming video-to-video generative AI app release.

$583M: Ashton Kutcher’s AI venture capital fund. And the raise happened incredibly fast, hitting its goal in less than five weeks and is set up to fund “category-defining artificial intelligence businesses at the foundation model layer.” The fund portfolio includes Anthropic, OpenAI, and StabilityAI.


🤖 Dub your video content with Paper Cup. The synthetic voices sound human (and you likely won’t be able to tell the difference even if you listen closely).

This platform, trusted by Bloomberg (to name one), integrates what you create and distributes multi-language dubbed videos across social platforms, allowing you to validate and reach a larger audience with “low risk, light effort.”

🤖 Runway is a full AI tool library with a tagline of: “Everything you need to make anything you want.”

Bold statement, but true. I’ve been using it for a while now and it can produce just about anything you want. It is a 30+ AI creative suite that is taking creative teams to unparalleled dimensions in their marketing with “No lights. No camera. All action.”

How well do all these AI tools work?

This 45-minute breakdown from Alex Hormozi sharing how his team has used some of the above and other AI plug-ins and products to keep up with his content demand.

It's a great example of harnessing the power of AI for acceleration and long-term growth using teamwork.


How will AI affect traffic, across SEO, SEM, paid ads, newsletters, and so on?

I have an idea—and so do many others.

And we’re speaking at an event next week that you can attend in-person or via Livestream (paid).

Normally, I don’t pitch anything in these emails.

But this event, by Perry Marshall (the guy who cracked the code on Google PPC), is too important to ignore:

I’m speaking in two sessions at the event (and that’s an affiliate link above).

If you want the non-affiliate link, click here.

Some of the world’s best traffic experts will show you how to get more leads, sales and moat-building data… with Artificial Intelligence, Facebook, Youtube, Google Ads, TikTok and more

My talk is Bionic Analysis: How to use AI and ChatGPT to optimize anything for more clicks and conversions.

I’ll be sharing some things I haven’t revealed anywhere else (not even in my AI copywriting workshop).

If you join us via my affiliate link here, send me a copy of your receipt, and in June (date TBD), I will custom write prompts for you and your business.

How many prompts? Depends on what you want them for, but you can get the best outputs for anything with up to 5 or 6 of them. I’ll give you all the prompts you need, according to my years of writing prompts and using AI.

And these prompts work on any language model and will never stop working.


Also, in case you missed it, a special issue on AutoGPT for marketing dropped on Monday.

Next special issue will be on the Code Interpreter for ChatGPT. You don’t want to miss it. It’s a bigger deal than Plugins.

See you next issue,
Sam Woods