Friday Long Read: What To Do About AI
This is a Friday long-read, so grab a warm cup of something and kick back because we’re going to take our time on this. The world is about to profoundly change. I know you’re nervous – perhaps excited and optimistic, but if you’ve been paying attention and have been watching the trajectory of this thing, the rational reaction is to be nervous. In this post I’d like to unpack in practical and tangible terms what AI is, where it came from, and the state of play, and then I’ll show you a path that will give you a pretty good shot at surviving the coming revolution.
Who am I? I’m Mark Maunder and I founded Wordfence in 2011 and wrote the early versions of the product until 2015 when Matt Barry took over as lead developer and I morphed into a tech executive running Defiant Inc, which makes Wordfence. We have over 4 million customers using our free product and a large number of paying customers using our various paid WordPress security products. I’ve been a technologist since the early 90s and a kid hacker in my teens in the 80s and 90s. I started my career in mission-critical operations for companies like Coca-Cola, Credit Suisse (now UBS), and DeBeers and then went over to do dev for companies like the BBC and eToys which was one of the biggest dot-com busts. I created the first job meta-search called WorkZoo in the UK around 2001 which later competed with Indeed, launched after us, and which I sold in 2005 but which made Time Magazine and NYTimes. I subsequently launched a Geoblogging platform, inline comments via JS, an ad network, real-time analytics, a localized news website, and more. You could say I’ve been in the innovation game for a minute, and I’m in it for life. I’m based in the USA these days in case you’re curious – moved over here in 2003.
My apologies for the long bio, but what I’d like to illustrate is that I’ve seen tech come and go. The hype cycles I’ve seen typically include:
- Outlandish claims about how the tech will solve everything from slicing bread to world peace and everything in between.
- Commercial vendors jumping on the emerging mega-trend to surf the wave with proprietary technologies of their own which they position as standards, or at the very least the default choice.
- Nascent technologies implementing the tech, that are immature, unstable, rapidly changing, and may very well be abandoned in a few months or a few years.
- The press contracting a bad case of rabies and foaming at the mouth about the tech, amplifying the most extreme aspects and use cases and creating a lot of noise, which makes it hard for implementors to sift for the truth and the fundamentals around the technology.
- The investment community pouring cash into the space with little focus. This creates an extremely adversarial environment for tech practitioners who are building fundamental value, who now have to compete with powerful VC-backed marketing machines.
- As Warren Buffet says, the Innovators, the Imitators, and the Swarming Incompetents enter the space in that order. I’d add that they have a pyramid structure with each successive wave being at least an order of magnitude bigger than the last. Things get crowded for a while.
- Then you have the typical bust cycle which cleans house and makes the tech uncool again, but also makes it interesting to the true believers. The VC’s go away and stop making noise that innovators have to compete against. The imitators and swarming incompetents drift off to imitate and mess up something else. The businesses not creating fundamental value fail. Some creating fundamental value fail too but the talent and tech are sometimes reincarnated into something else useful.
So who prospers, and what tech survives after a bust? Sometimes none of it, but helpful derivative technologies are created, like Java Applets in the 90s that inspired Flash which inspired standards-based rich content web browsers.
Sometimes out of the ashes, an Amazon is born, as with the dot-com boom. And sometimes you have incredible innovation where the innovators never see large-scale commercial success, but others do, as with Igor Sysoyev who created Nginx which eliminated the need for a data center full of web servers to handle large-scale websites. Igor has a commercial thing, but the real winners were companies like Cloudflare who based their global infrastructure on Nginx, reverse proxying massive numbers of connections to origin servers with rules about what gets proxied. Hey, I benefited too. Nginx saved our behinds when Kerry and I were running Feedjit.com from 2007 to 2011 because it let us handle over 1 billion application requests per month on just 6 servers. Thanks Igor!
Blockchain technology is in a bust cycle and you can map the characteristics I defined above to blockchain. It looks eerily similar to the dot-com bust and you’ll see an Amazon emerge from the ashes about a decade from now. It might be Coinbase if they survive the over 80% dive in market cap that may continue, but who knows?
Derivative Versus Fundamental Technology Innovation
I’ve mentioned a few tech bubbles so far and it’s really the first step in pulling us out of the trees so that we can examine the various forests out there. Now let’s go a little more meta and talk about which forests matter. Some technology is derivative. Examples of derivative tech are new stuff that runs inside a browser, for example, Websockets.
Websockets are awesome because they let a browser keep a connection open and get push notifications without doing the old TCP three-way handshake to establish a new connection every time the browser wants to check if there’s data waiting on a chat server or whatever. We used to call this long-polling and I wrote a web server to do long polling which was a clumsy but necessary approach, so when Web Sockets came along we all breathed a sigh of relief and I happily retired my web server glad that no one would see my nasty source code which worked quite well mind you.
Another derivative technology – and you’re not going to like this – is blockchain. It’s a useful and novel implementation of hashing algorithms and a few other cryptographic tricks, but honestly, we should have disintermediated banking at least two decades ago and the fact that blockchain has still failed to do that is both disappointing and illustrative that it is just a set of derivative applications built on a tower of fundamentals that has a way to go before it matures. The hype cycle and speculative bubble around it was simply humans making human noise.
So that’s derivative tech. In addition to derivative tech, you have what I’d like to call fundamental tech. Electricity is fundamental. It’s cornerstone technology that transformed the world in our ability to use and transport energy which has enabled an industrial and technological revolution the likes of which the world has never seen. The microprocessor is also fundamental tech for similar reasons. You have algorithms that are fundamental tech like the RSA algorithm which allows us to establish a secure communication channel while a bad person is listening in the whole time – the kind of tech that could have changed the outcome of World War II.
The Internet is fundamental tech. It connected the world and gave us the ability to build applications on top like HTTP and the web which are derivative tech.
Oh I know you want to have a bar fight with me at this point and we’ll do that if you’re attending Wordcamp EU – a collegial and metaphorical barfight, that is – but hopefully, you’re picking up what I’m putting down here in a general sense: There is fundamental tech that profoundly enables and changes the world and which many other things are built on top of, and there is derivative tech that gets a lot of attention but isn’t quite as transformative in a historical sense if you’re thinking in terms of centuries. And there’s the big fat grey area in between.
AI is fundamental tech. For decades we have been programming by writing functions by hand. We’ve gotten quite good at structuring our code using metaphors like object-oriented programming to create logical structures that make sense in a human world, and help us organize large code bases. But fundamentally the way we define logic in a program hasn’t changed for a long time. Until now. For the first time in all of history, we can create functions in programming by training them, rather than writing them by hand. In other, slightly more technical words, we can infer a function from observations and then use it. Like babies and toddlers do. This is historic, it’s transformative and it is a fundamental breakthrough.
Funny thing is that until quite recently – around 2015 – AI had suffered many so-called “AI Winters” where there was significant interest in the field that catalyzed investment dollars, and then a setback usually caused by a reality check, that caused a winter in funding and interest. Does anyone remember the “expert systems” of the 80s? By the early 2000s AIs name had been dragged through the mud so many times that anyone doing serious research in the field used different words to describe their work, like “machine learning” or “informatics” or “knowledge systems”.
A few hardcore true believers like Yann Lecun, Yushua Bengio and Geoffrey Hinton powered through like Bilbo and Sam across The Dead Marshes and went on to win the Turing Award, which is basically the Nobel Prize of computer science and which I had the privilege of attending when Rivest, Adleman and Shamir won theirs for public key crypto. The Turing Award is a very big deal and well deserved considering how adversarial the AI environment was for a while.
So what changed? Well for one thing you’re reading this post because it’s about AI and you’re interested. And you’re interested because you recently used GPT-4, MidJourney, Dall-E or another model to create something. You’re seeing tangible results. And the reason you’re seeing results is that GPU hardware, algorithms, and an interest in the field have brought us to an inflection point where the technology is delivering results that are jaw-dropping enough to catalyze more funding, more research, and more jaw-dropping results. This cycle really picked up steam in 2015, and with the release of GPT-4 recently, has entered a phase of what I would describe as true and consistent exponential growth.
According to NVidia “LLM sizes have been increasing 10X every year for the last few years”. In two years that’s 100X. Three years from now that’s 1000X and so on. Extrapolate that out and be afraid. Or optimistic if your mind isn’t for rent and you are hopeful yet discontent. Rush lyrics aside, that pace should give you an idea of how quickly this thing is coming. And now that we’ve reached the point of inflection I mentioned, where the hardware and algorithms seem to have overcome the cycle of disappointment that AI has been stuck in for decades, I predict that you’ll see consistent and exponential growth in the field in capabilities for the foreseeable future, with a financial bubble and bust in there that won’t be of much consequence to the fundamental value of the technology.
“Thanks for the history lesson Maunder, but you brought us here with promises of telling us what to do about AI. So?”
What to do about AI
So far we’ve discussed what boom cycles look like and the kind of noise and bear traps you should be aware of. We’ve defined what AI is in fundamental terms – a function that you can train rather than hand code. And we’ve hopefully agreed that we’ve entered a period of consistent and exponential growth in the field. Now we’ll chat about how to survive and prosper in a world that looks a lot like when electricity was invented and commercialized, or the microprocessor, or the Internet.
Goldman estimates that AI will add 7% to global GDP at a rate of about 1.5% growth per year. They also estimate that roughly two-thirds of US occupations are exposed to some degree of automation by AI. You can extrapolate this globally. That kind of global disruption is matched only by the industrial revolution or the entire recent tech revolution as a whole starting from 1980. From the same publication, “A recent study by economist David Autor cited in the report found that 60% of today’s workers are employed in occupations that didn’t exist in 1940.”. So on an optimistic note, this kind of disruption isn’t a new thing and we’ve been disrupting and adapting for some time now.
Perhaps you’re reading this because you are running a WordPress website, perhaps secured by my product, Wordfence. Which means you’re a creator of some kind. Perhaps you’re a writer, an artist, or perhaps you’re an entrepreneur creating a business out of thin air. [Yes my fellow entreps, you get to hang with the other cool creator kids too!!]. If you don’t plan on adapting at all, that makes you far more vulnerable to this coming wave than say a chef who runs a restaurant, or someone who manages real estate and rentals. And that really is the key: adaptation. So how can we adapt?
If you’re a creator, you need to become a user of AI. You’re probably already using GPT to write copy for your product catalogs on your e-commerce website, or using MidJourney (MJ) or Dall-E to create art for ad campaigns. If you’re a designer or artist, you may feel the kind of resentment this Blender artist does in the Blender subreddit.
“My Job is different now since Midjourney v5 came out last week. I am not an artist anymore, nor a 3D artist. Rn all I do is prompting, photoshopping and implementing good looking pictures. The reason I went to be a 3D artist in the first place is gone. I wanted to create form In 3D space, sculpt, create. With my own creativity. With my own hands.”
“It came over night for me. I had no choice. And my boss also had no choice. I am now able to create, rig and animate a character thats spit out from MJ in 2-3 days. Before, it took us several weeks in 3D. The difference is: I care, he does not. For my boss its just a huge time/money saver.”
While I sympathize with how hard change and disruption can be, it’s been a constant for the past couple of centuries in many fields. MidJourney has a long way to go before it can match a real-world artist, unless you’re just churning out images and letting the AI guide the design choices and are happy to work around the bugs. For MidJourney and other generative AIs to produce exactly what we want, they’re going to have to get better at understanding what exactly we want to create. And that’s where the skill comes in. You’re already seeing this with a document that someone has created listing famous photographers and examples of their look. This can be used in MJ prompts to say “in the style of” to get a specific look, but it is an incredibly rudimentary approach.
Another way to guide the MJ AI in particular is to blend photos it has generated. Again, super rudimentary, but it’s the start of having the ability to tightly specify exactly what you want and get that out of MJ. And if you need a reminder of how basic it still is, try to get MJ to generate hands. It still sucks, even at version 5.
So if you’re a creator, start getting good at using the tools now, understand their limitations, and evolve as the products evolve until you’re an expert at guiding the AI to create exactly what you want. This will help you guide your customers in explaining the limits of the current state of AI to them and where you add value, let you take immediate advantage of the use that the current tools have, and ramp up your productivity as the tools get better at taking instructions from you.
This applies to writers, artists, designers, filmmakers, photographers, screenwriters, and anyone with creative output. Get good at the tools. Get good at them now. Do it with an open mind. Know that changes aren’t permanent and that change is. (Again with sneaking in the Rush lyrics)
Adapting as a Dev
Coders! My people! We have a problem. Most of you have become users of AI. You’re users of GPT-4 via their API. You’re plugging into other generative AIs via an API. You aren’t rolling your own. And rolling your own is where all the fun is!!
Ever heard of transfer learning? You can grab a pre-trained model from Hugging Face, chop off the head – aka the final layer in the layers of neural nets, substitute it with random weights, and train the pre-trained model with your own data to take advantage of the sometimes millions of dollars that someone else already spent training their model. In fact Facebook’s LLAMA model which is one of the largest LLM’s in the world was leaked via Torrent recently.
The most important thing you need to do right now as a developer is to stop being a user of AI and become a dev of AI. GPT-4 is a shiny ball that the world will have forgotten about in a year, but it’s a very shiny and attractive ball right now that is fueling many a late-night dev chat. Remember that stat I gave you above? That LLM’s have been increasing in size at 10X per year. The current state of the art will be accessible to you on a desktop in a few years and you need to get ready for that world today.
I’m going to just go ahead and tell you what you need to do to get your AI stuff together, fast.
- Ignore the math. Trust me on this. Most people including devs are not good at math and it intimidates the hell out of them. AI is just matrix multiplication and addition using GPU cores to parallelize the ops. Expressing this as code is easy. Expressing it as math will make you hide under your bed and cry. Ignore the math. If you can code, you’ll get it.
- Learn Python. Everything in AI is Python. It’s a beautiful little language that you’ll come to appreciate very quickly if you’re already a dev. It’s like coming over to Aikido if you’re already a black belt. OK the MMA scene kinda messed up my metaphor proving that Aikido is actually worthless, but whatever.
- Then go do the Practical Deep Learning for Coders course at fast.ai. It’s how we get our guys up to speed fast in the field and it’s brilliant. Jeremy Howard does a spectacular job of getting you up to speed fast in the field by immediately getting you productive and then unpacking the details in a fun non-mathy way.
- As you progress in the course, definitely get up to speed using Jupyter Notebooks and I’d recommend Kaggle for this. They were bought by Google a few years ago and kind of compete with Google’s own notebook system called Colab, but I prefer Kaggle. You get GPU access by simply verifying your email address and it’s free which is kind of amazing. So you can use a rich text environment on Kaggle to write your code, see the output and run it on some fairly decent GPUs. Kaggle GPU’s perform at about 20% of the speed of my laptop RTX 4090 in case you’re curious about benchmarks.
The course teaches fundamentals, how to use pre-trained models, how to create Jupyter Notebooks or fork others, how to create Hugging Face Spaces, and how to share your models and their output with the world. It is the fastest way right now to transform yourself from an AI user into an AI dev and get drinks bought for you at parties by folks that have not yet made the leap.
Alright, this went long but that was the plan. We’ll talk more about AI. Go forth, be brave, learn, and create!
Mark Maunder – Founder & CEO – Wordfence and Defiant Inc.
Footnotes: All images on the page were created with MidJourney and if you’d like to see the prompt I used, simply view the image in a new tab and the image name is the prompt, all except for the heavy metal hands image which a colleague created. I’ll be in the comments in case there’s discussion.
The article discusses the potential impact of artificial intelligence (AI) on society and offers practical advice for navigating the coming changes. The author distinguishes between fundamental technologies, such as electricity and the internet, and derivative technologies, such as blockchain and websockets. He argues that while derivative technologies can be useful, they are not as transformative as fundamental technologies. The author also discusses the hype cycle that often accompanies new technologies, and the importance of focusing on the fundamentals rather than getting caught up in the hype. Overall, the article encourages readers to approach AI with caution, but also with an eye towards the potential benefits it could bring.
Ha! I'm guessing you pasted the full article into GPT-4 and asked for a summary. If anyone else posts anything generated please do include the source to satisfy the rest of our collective curiosity. Thanks.
Wow, thank you Mark. It would be a bit of understatement to say you just handed me the gift of a lifetime. With gratitude, with appreciation, B.
You're most welcome Bodhi!
I tried to get a summary for my readers from chatgptplus/4 and got the following error:
"The message you submitted was too long, please reload the conversation and submit something shorter."
Just thought it was interesting to share.
Thanks always for the work you do.
You’re very welcome! Yeah increasing the token limit on large language models is going to be a big enabler.
As a creator, where would one start in order to "get good at the tools"?
Hi Lori - thanks for the question. OpenAI is currently the state of the art in NLP's (natural language processing) models with GPT-4 so start there and I'd strongly recommend getting a paid account to get access to GPT-4 which is a significant step up from GPT-3.5. It'll really give you a feel for where the tech is at, and it's pretty much shocked the industry in many ways.
Also, their image generative AI called DALL-E is pretty good: https://openai.com/product/dall-e-2
MidJourney is in my opinion the state-of-the-art in generative AI's that produce images. They're significantly better than Adobe's Firefly which is a bit of a joke, to be honest. You'll need to install Discord to interact with MidJourney but it's worth it and I have a paid account for private fast generation which I've found to be worth the rather expensive $500'ish per year.
And Firely is here: https://www.adobe.com/sensei/generative-ai/firefly.html
I'd suggest starting with those and get good at understanding what a prompt is, how to craft a prompt (I include some comments on this in the article) and what's possible and what may be possible in future.
First - Thank you for an enjoyable read, all the way through! I wouldn't discard Firefly just yet, and it truly isn't here yet. For Adobe users, it will integrate flawlessly with the other Adobe products, as always, and it will provide a different interactive experience between the user and Firefly as it will not only tap into its own library of stock imagery but also my own Adobe library, right? Or did I drink the Adobe Kool-Aid?
Haha maybe. It's interesting how the images generated by Firefly seem to have a strong phobia towards copyright. So much so that it produces mush if you ask it for things like images about Deadpool, or Spongebob. At least in the samples I've seen.
Brilliant and insightful article. Thanks for creating this. I thought the overview of the boom-bust cycles of tech was illuminating. I found that I was still a little muddled about derivative vs fundamental, as you "defined by example." I had to dig a bit to understand how you were discerning between the two, and what was in the big grey area. I'd write more, but I have to head over to fast.ai now...
I'm really glad I've sent you over to fast.ai Mike. It's pretty much where I want everyone who wants to learn AI deeply to go immediately after reading this. Every minute you spend there will pay off well over the coming decades, and score you serious points with friends/family who have questions. The field has become a lightning rod for ethicists and doomsday prophets leaving little room for gaining a fundamental understanding of it all so that one can form one's own opinions.
How about AI for websites, how will it impact let's say SEO or SEA or online advertising in general?
Great question. So the biggest change is that Google is dead. The idea of asking a search engine a question that takes you to someone else's website where they can have their way with you is going to seem quaint in a few years. Google has launched many products but fundamentally a huge chunk of their biz relies on what John Batelle called a "database of intentions" about 20 years ago. The power of Google is that they know what you want when you want it and can serve you targeted ads at your moment of need. They only reason they have this privilege is that they dominate search and have for over 20 years. If you haven't tried GPT-4, try using it as your search engine for a bit and see how it changes your behavior. Having a concise textual answer is what many people want much of the time.
Yes Google is working on it's own models, but it would have to cannibalize it's search business to launch them and dominate the AI space, which it won't do.
A buddy who works at Microsoft was telling me that Bing's AI integration and some of the AI integration in Edge is pretty damn impressive. I haven't looked at it myself yet, but it makes sense since MS invested about (reportedly) $10B in OpenAI and has full access to all their intellectual property. So of course they've integrated it and of course it's the core of their strategy to take out Google.
So I think these tectonic industry shifts will have a huge impact on SEO and how we promote websites. It's quite probably that we'll see a website collapse (peak websites?) as generative AI models train on our content and then produce their own answers without citing sources and without sending the asker to the source. It may create a very serious walled garden AI crisis.
Online advertising will move to the purveyors of AI because that's where the eyeballs are along with where the database of intentions has moved to.
Big changes a comin!
Thank you Mark, for your reply. So you are saying SEO-specialists and affiliate marketeers should focus on something else then creating websites?
That’s discussion for another day. I’ll look at doing a follow up.
Hi Mark, very interesting post. I think something is missing after Time Magazine and NYTimes. “I created the first job meta-search called WorkZoo in the UK around 2001 which later competed with Indeed, launched after us, and which I sold in 2005 but which made Time Magazine and NYTimes.”
Not really - you mean it seems like a typo? Or you want the rest of the story? Sorry not sure what you mean. I guess to fill in more detail, Kerry and I moved to the USA from the UK, relaunched it here, it became very popular because Google had just IPO'd and "vertical search" was the new hot thing, and we got a ton of press, got an offer and decided to sell. I worked for the buyer for 8 months and then dove straight back into full-time entrepreneurship based in Seattle. Lots more to tell I guess but out of the scope of this article.
Sorry, I misread a word in that sentence. Everything is good. Thank you for the extra bit of information and for sharing your thoughts about AI.
You're most welcome Arturo!
Hi Mark, thanks for your post. I've just finished the book, The Age of AI and Our Human Future by Daniel Huttenlocher, Henry Kissinger, and Eric Schmidt. I recommend it as a thoughtful -- sometimes hopeful and sometimes frightening --- look at our future with AI. I recommend it to supplement the information you've provided in your article.
Thanks for the extensive, detailed and long note. I enjoyed it for what I am, a 'mutant child' of the marriage between art and technology.
Thanks Jorge! We're all mutants of a sort.
Wow, thanks for this. Fascinating article. Took me an hour and a half to read it because I kept going down various rabbit holes you pointed at. Super exciting times.
You’re very welcome! Glad to hear you went exploring!
This article was informative, and to the point... and it was long too... but it's Friday too.
AI is more than exponential, and as it grows, and the LLM gets bigger and bigger, I wonder what goes by the mind unnoticed without discovery, and then errors go into exponential growth mode too... we'll need AI to solve the errors, but the AI made the errors in the first place.
So it brings up another dilemma... we set AI to fix what AI created in the first place, and who knows what evils lurks in the heart of man, the creator of AI?... my guess will be God
As a creator, I must say, the future you paint is bleak, because creation is what makes life worth living. And before you say it, AI does not create, it appropriates! I've been at it for 70 years and have loved every second, even moving from the airbrush to a computer (though I doubt you know what an airbrush is or care, though I see AI ripping off that style all the time). But even so, the computer was still just a tool to be harnessed. It still required talent and vision. The fact that any schlub can "create" with AI, (a truly “derivative technology that preys and can only exist on the creativity of actual artists) and that we welcome this, just demonstrates how bankrupt we are as a society. I feel bad for my grandchildren, and all the artists, musicians, writers that will slowly vanish over time, and there will be no bringing them back, the true visionaries will be gone forever, and we’ll be left with the greedy and the money hungry, leveraging AI for every last f****ng penny! But young people are so hypnotized by technology that they don't get it; as you said, "Know that changes aren’t permanent and that change is." You quote the very artists that will be lost forever! How does that work? I use Wordfence and have loved it, but as my importance as an artist and creator vanishes, so will my websites, and so will my need for Wordfence and my web host and my computers. There are consequences to putting machines over humans, but we remain as stupid as ever—society devolving into chaos, people with no regard for one another; we deserve the future we are creating, though I'm not sure our children do. Call me a Luddite. But think about this; why don’t we remove all speed limits on our highways, and take down the traffic lights and stop signs? Because it’s dangerous and people will die; luckily sober minds have prevailed in those areas and realized to do such a thing is insane. We don’t have sober minds when it comes to money and greed, and that’s what AI is, another fast track to wealth and riches by the lazy and uninspired, sociopaths so spiritually insolvent that no amount of money will ever fill the void. Progress and technology, machines over humans, like speed limits and traffic lights, are choices, not eventualities. I know that technology has cured cancer, eradicated poverty, abolished social and racial unrest, has destroyed income inequality, and thank god, has finally solved climate change! Or not.
I’ve included this comment to add a complete perspective after removing the f-bomb. Other than that it is unchanged. To answer your implied question: I’ve used an airbrush.
I was hoping to learn more about any security risks that we might encounter as a result of AI - especially as it pertains to WordPress. I created a social media platform. If you haven't, then you might be surprised at how difficult it is to get cyber breach insurance, even with Wordfence Care installed. In fact, I'm still searching.
The main concern is that my users can upload their own content (UGC). This seems to freak out the insurance companies. Users on my site can create AI-generated blog posts or upload other AI-generated content.
Do I need to be concerned? What precautions can I take? How is Wordfence addressing these concerns?
I'm grateful for WordPress, BuddyBoss, Elementor, and Wordfence that allow me to build upon the knowledge and skillsets they've provided so that I don't have to be a coder. But, I do have concerns about the AI bad players. Security and safety are my MAIN CONCERNS for my WordPress site. I'm counting of Wordfence to keep my content and my users' content safe.
Thanks Dianne. More to come where we’ll address some of the risks you’ve asked about. In a reply to another comment here I do discuss some of the changes in the strategic landscape that I see coming.
Hello, great post thank you. As a transforming technology, beside SEO (that may become 'the old world') and infosec, how will AI impact WordPress website designers, and WordPress in general? How do you imagine the impact on such 'old' technology? I would love to know how you see the future of WP in the AI era!
Ho, and any advice for a WP site designer and/or plugin dev would be much appreciated -:)
Hi Mehdi, see my other comments here where I provide additional answers and resources. Let us know if you have a specific question.
Hi Mehdi, sorry I've been replying to comments out of order. Much to say about this and I've said a bit here already, but thanks for your question and I may do a follow-up piece.
Thank you Mark! Looking forward to read you once again then.
Hi Mark. Thanks so much for this informative article and for inventing Wordfence. Great product. I’m a WordPress web designer who mostly works with themes, web builders and tools like Advanced Custom Fields to produce sites. I’ve been playing with GPT4 to help shortcut research time via Google or YouTube when troubleshooting, say, an integration issue between ACF and Elementor, only to be given outdated information. I keep trying to use the tool because I do feel like it’s helping me to save time finding out what doesn’t work. And it can be fun to use. But I’m wondering when the Sept 2021 data threshold is going to be lifted, and if that will even solve this issue?
Yes Mary great point that I didn't include in the article - the cutoff for training data on models like the GPT-4 LLM. I have a feeling it's not just that training takes time. GPT-4 has some pretty significant controls on what you can ask it and having non-current data may be one of those controls. Just a theory.
Indeed, this is mindblowing. I’ve been able to create marketing content in seconds rather than the half-hour it used to take me to craft an effective email for the work I do in sales.
However, what scares me is AI’s ability to “scoop” available content. I’ve had a WordPress site for the past 15 year crammed with articles that serve as a storehouse for categorized training in the field of Advancement for schools. In doing metaresearch to discover if any of the constructs I’ve written about have been utilized by other service providers, I’ve discovered I’ve been able to thwart copying potential by installing a No Right Click plug-in.
The next step is to make access only available by membership since there are ways to get around the no right click issue.
My concern is that if my content is protected via a membership, is there still the potential for one of those members to have my content be scooped since they would have access to it? Perhaps those protections don’t even matter if the escalation of AI ability continues to be exponential and can pull from countless amounts of resources - whether they’re scientific, truthful, opinion or conjecture.
Yes I'd say the finest real-world illustration of the debate about whether training models on copyrighted works is OK is this proposed class action lawsuit: Github devs versus MS/OpenAI/Github who scraped their code to train the copilot model:
I'm wondering if you train AI on open-source code, does the output carry the open-source license restrictions? The core of open-source licensing is to ensure that derivative works can be freely shared. To what degree might the output of AI be bound by such a license, which could force the developer who incorporated code from the AI to be bound by an open-source license. It seems like it could get rather thorny.
I am really glad that I clicked the link from Wordfence newsletter and read this article.
Thank you Mark!
I am wondering, what would be the next stage when a critical mass of creators will be using AI to create content or artwork for websites?
I guess that this AI created work will have a minimal value and will generate minimum profit for these creators. AI created content and artwork, will be like a mouse trap for them.
Your source to https://course.fast.ai/ is priceless!
Yes Ilias I agree, the market will shift as will pricing and what is valued. When everyone has access to generative AI and can create high quality images, where are the opportunities to add value? I think that's exactly what I'm trying to get at in the post regarding how designers should adapt: Find that value and be the first and best to market.
If I had to summarize your article in a few a words, it would have been something like this: "AI is powerfull tool and if you want to stay competitive you'll have to learn how to use it, or how to maintain, modify or enhance it".
This applies to any other tool ever invented. Your text clarifies that AI is another great tool.
Thanks for the effort in putting together this great insight and sharing with the world.
You're welcome Nebojsa!
Humans making human noise… Love it.
Superb comment Lonnie Busch. I agree 100%. What a soulless world we are creating.
Hi, Mark. Your article was just what i needed to get started with AI development. I have already developed few stuff in python and i really love this language, so i thought about getting into AI dev, but most of the courses do recommend a strong math abilities, so i had my doubts about my ability to participate in development. Yet when it's coming from someone like you, this article has inspired me to try and that's what i needed really. I thank you from the bottom of my heart, even if i don't succeed at least i will try.
You're most welcome Avi. Glad I could help and inspire.
Will AI be able to produce high quality travel and landscape photographer as if they were there ?
Lets find out. Heading over to MidJourney to test. I'll be back momentarily. It's 4:43pm mountain time.
I'm back at 4:49pm 6 mins later and here's what I made. David you may have found kryptonite for generative image AI's. This is "A view of Table Mountain from Milnerton beach in Cape Town with a cloud rolling over the mountain and a stormy sea." which is a famous view in Cape Town and it's where I grew up. If you've never been there the view kinda looks OK, but it's profoundly wrong in subtle ways, mostly to do with the actual topography. If you live there you'd immediately recognize this as fake and I think it would be very difficult for an AI to train up enough to get topography in photos accurate - hmmmm actually now that I'm thinking about this I might disagree with myself. It's just data expressed as 3d coordinates and that's the kind of data that NN models love to gobble up. The photo realism is off but I may have just had to add 'photo realistic' to the prompt. Anyway here's the current state of play from MidJourney:
Thanks for doing that really appreciate it. I'm a professional photographer (delveintoeurope.com) sorry not sure if that is allowed but wanted to show you the type of photos I do to compare the quality of above. I'm also a writer and I'm aware AI will have a massive affect on writing but was curious about photography. Just wanting to figure out what I should focus on, where I can add most value and what I should be doing to adapt.
No problem. Nice site. I've made the link clickable. If you're going to WCEU2023 I'll be there and we can say hi! I'm a 6ft1 redhead so kinda hard to miss.
Thanks Mark, that's very kind. I'm 6ft3 lol. Do you think AI will replace photography like mine ? Or sites like mine will be obsolete in a couple of years ? Any tips on what I should do to keep ahead in these crazy times.
It won't. For example there's an opportunity around creating generated data from your own training data even if it's one sample. Think of GPT-4 when you give it the blog post above and ask it to summarize. (It's a bit long actually but you get the idea) So there's an opportunity to provide your image or images to a model and get it to generate stuff. Just always beware of copyright - which means you'll probably have to pay for this capability to ensure it stays in your own little walled garden and doesn't get used to train a model that others will use to generate images like yours. To understand opportunities like these it helps to understand how the tech works so that you have a vocabulary and can ask the right questions of a human, AI or search engine to figure out a path forward.
Thanks this is really helpful. I'm 50 plus and originally trained on medium format film cameras ! so seen a lot of changes and feel like a bit of a dinosaur when it comes to this stuff. Great example of providing it with images but making it stay in my own garden. That was my fear that AI could just scrape my photos and others could start using them. You are right on need to learn how this tech works. Thanks again for taking the time out to answer my questions, really appreciate it.
Enjoyable read, Mark. I haven’t played too much yet so the AI stuff was informative, but everything else in the article I knew (I’m very old) but hadn’t ever seen laid out so coherently. Off to register a paid GPT account and get my feet wet, and the whole will be required reading for my developers.
First off, I appreciate your intro, because I also have been involved in tech in various ways for a long time, watching patterns such as those you describe unfold. But with about a decade head start, before the internet was significantly "public," a bit before the WWW even existed, and before it was all commercialized. And, my bias is in favor of humans before/over technology, so we differ there at least.
Second off, triple dittos and seconds and thirds to the point made by Lonnie Busch that AI is essentially an appropriator of human work and human creativity, with no credit or pay to the workers/creators. Less tactfully: the builders of AI are ripping off humanity's collective creativity over many centuries for private profit for a tiny number of humans, most of whom are already wealthy beyond all need or reason.
The appropriation is beyond obvious. What is AI trained on/with? The works of humans. Without massive, unpaid, unacknowledged consumption of centuries or even millennia of previous human creativity, AI would be...what? Hard to tell. Maybe some AI theorists somewhere are playing with those kinds of models, but it's not going to make money for VCs so it's irrelevant.
Essential additional factor: AI is built almost entirely by privileged white male humans and trained predominantly on/with works of dead white male humans, thus no surprise AI has already demonstrated the same white-supremacist and misogynist (among other nasty) behaviors as facial recognition technology has previously shown. Anyone not working to correct that with all available resources is implicated in the consequences.
Third off - nothing to do with the post content - it's beyond small-world spooky that I just happened to decide to read this on a whim, and found that two of the first three comments are from people I actually know! Hey Bodhi and Brent, howzit?
Thanks for your thoughts John!
My initial thoughts.
* I expect legislators soon pushing for laws to require stating when content is AI authored/co-authored, and to dictate a "bibliography" of the reference articles/images.
* I'm a techie not copywriter, but I'd expect for the moment good ones'll remain relevant by being able to not just string coherent copy together, but to know the best copy to speak best to their target audience, be in slang, local/cultural references, etc.
And heaven forbid when politicians get on board with it - what they say is already oft detached from the reality of their beliefs :P
As popular as it has become, and also that rapidly, AI is generally presented as "a technology." Is that fair and accurate or is AI an umbrella term that encompasses a variety of technologies?
Hi Cemal. Fundamentally it’s a bad moniker for layers of neural nets that are trained using labeled training data. The trained model is then evaluated, tuned and then used as a function with inputs and outputs. An example of this is the labeled data that Cambridge University has created for self driving cars. They’ve labeled 20 minutes of video from a feed from a car driving through a town. Pedestrians are say purple, cyclists are green, stop signs are red etc. That labeled data can be used to train a model that will then be able to recognize fresh video and which can be used by a self driving vehicle to make navigation decisions. Deep learning is just another name for layered neural nets. Most people who write about AI don’t know what it is or how it works. So it ends up being portrayed as a black box that does scary stuff. It’s not that hard, not that scary and is very powerful tech. I think part of my goal with this post is to provide people with tools to understand AI at a fundamental level rather than just as an abstraction. So I’d say it’s not really an umbrella term, it’s quite specific, but the range of applications and implementations is huge. Hope that helps.
Thank you. Although I don't fully understand the intricacies, I have some idea of AI. I looked at AI in the context of photography in a recent article. There, I focused mainly on the characteristics of photography rather than how AI works.
I also have a hangup in my mind that feelings and emotions are a part of "intelligence" which guides how human minds and intelligence process information. Any thoughts on that?
Great read. I am about 4 weeks into my AI deep dive for understanding the basics and current status of the prompts, language models and text to image generator hubs like BlueWillow etc.
Qs: Do you think website template portals and hubs see the writing on the wall if AI can spit put a custom themed WordPress or Joomla custom site based on a detailed prompt?
I think we’re a couple years away but yeah it’s possible. And by “couple” I really do mean 2.
Human action originates from two fundamental sources—cleverness and wisdom.
I believe AI is going to be the ultimate litmus test for determining whether or not humans can choose wisdom over cleverness.
The wrong choice will mean the end for mankind.
One of the great flaws of humanity is the this, and it’s quite simple….
“Because we can we should.”
The use of AI for content creation will cause a disruption in the publishing and art world (especially writing and graphic design) and that is unfortunate. However, AI is also allowing others (like educators and freelance writers) to craft learning materials or article content they might not otherwise be able to create or afford to create.
As someone who works specifically on endangered species issues (journalism and education) I am currently exploring how to use AI to enhance the effectiveness of the content I create for all my projects. And creating effective content means creating compelling content.
I am not soulless or greedy, I am just concerned about the health of the planet and the other species we share it with.
As empathetic as I am to those who will be negatively impacted by AI, I am more concerned about the health and well-being of the other species we share the planet with. Especially those species in danger of extinction; endangered species.
The only way to help endangered species survive is to engage people to take action. Especially young people. And getting young people to take action first means getting their attention. And getting their attention means creating compelling content. And if you look at the artistic level of content that young people have access to, content created by big corporations like Disney or CNN, then you have to ask how a teacher or freelance journalist can compete for the attention of this audience?
I believe AI technologies like Open AI's DALL-E, Amazon's Polly and Adobe's Sensi (and other AI technologies) can help journalists and educators do just that.
As my first venture into using AI I created the video at the top of this page https://bagheera.com/malayan-sun-bear/ I used DALL-E for the images; Polly for the voice narration; Adobe Sensi for the music mix-down. It was my first attempt to use AI to create endangered species education content.
Since I am not an artist or musician (and since I don't have a wonderfully compelling voice), it would have been difficult for me to have created this content without the use of AI. I could have tried to find the images I needed from Adobe Stock or another content licensing business. Or I could have hired a graphic designer and a musician to create the art and music content I needed. I would have also had to hire a voice talent for the narration.
But the cost licensing content or hiring talent to create it could turn out to be extremely expensive. And that's assuming I could either license the content I needed or could find the talent to create it. And let's not forget that it takes time to find content online, and it takes time to find the talent to create it.
Then there are also the licensing and legal issues to be considered when licensing content from a business such as Adobe or hiring talent to create it. Both of these options require the signing of contracts regarding licensing rights. Who owns the content? How can it be used? How long can it be used? Where can it be used? And what if there is a misunderstanding regarding the contract; who pays the legal fees?
These are all very real questions that must be considered when either licensing content or hiring talent to create it.
However, using AI to create content removes all these issues. No licensing restrictions. No contracts. No time spent looking for content producers. And no legal headaches.
So, yes, I believe there is a place for AI in the world of content creation. Especially for journalists and educators.
AI, like any other technology is just a tool which can be used for good or evil.
Over the years I have had millions of learners visit my endangered species websites to get information about endangered species and the plight they face. ALL of these websites were built using modern technologies. So I am personally grateful for the technologies that allowed me to reach so many people with what I consider very important information.
My goal now, as Mark suggests, is to learn how best to use AI technology to craft content that might inspire young people to care about the planet they live on and the other species they share it with.
I am excited to see how AI can help me reach this goal.
MEd Learning Design and Technologies
Needless to say, I'm feeling overwhelmed but challenged. I dove headfirst into ChatGPT just a couple of months ago because I guessed that this was the way of the future, and if I wanted to be ahead of the crowd, I better get my feet wet now.
What adds to my confusion is the plethora of programs offering AI help with producing a boatload of results. I'm hesitate to spend money on something I think will be old hat tomorrow.
So for now, I keep working my way forward with chat.openai.com and read lots of articles. Like this one for example. :)
Thanks for the excellent insight, encouragement, and for the ongoing work of the team at Wordfence. Both are very much appreciated!
You’re very welcome!
I am a content creator, entrepreneur and lawyer so needless to say I have a number of connection points to this issue. It all really depends on what kind of content you are creating and who your target market is. As a content creator, I have been using tools like ChatGPT as helpful research assistants pulling together outlines of various articles I am looking on creating based on the framework generated by ChatGPT although the lack of sources is an issue. Andi helps in that regard. What I am seeing now is great basic information but often no context which is where I come in and is my value add.
As a lawyer, this is going to be very interesting to see how copyright plays out in all of this. Copyright does not disappear just because AI is being used so that is a challenge that the governments will have to confront ASAP. Does my copyrighted content appearing in a ChatGPT query constitute copyright infringement? I would argue it does but how is that enforced? It will be interesting to see how this plays out.
I agree there can, and will be copyright issues that arise with AI. However, I also think a user can be smart in the content they are creating, say for instance by not violating someone's "Right to Publicity" rights.
And there is the issue of my AI content being used in places I have not authorized. However I already have that issue with my current content which I create without the help of AI, which is content I have both the copyright and trademark for.
It is not an easy task trying to enforce copyright infringement. I mean have you seen how expensive it is to hire a competent copyright attorney : )
Your points about copyright are well taken. However I do believe the benefit of being able to create one's own content using AI, as long as it is done smartly, far outweighs the issues that arise in trying to prevent copyright infringement. Again, something I already have to consider with the content I create without the help of AI.
Not an artist, not a writer, not a coder, just an average IQ old scientist that, from the time I was a crawler sticking 2 keys in an electrical outlet, has been imprinted with the obsession of What!, Why, and How, thank you! Your article made me tingle like the melted keys that started it all. Now back to SEO, perhaps in a more optimistic way.
As far as what to do with AI in the immediate term, I think you have the right answer. There is literally no choice in many aspects but to leverage the technology.
However, if you play this forward, it becomes increasing difficult to reason about. There are already AI scientist who have made posts recently about burnout of just trying to keep up. I've call this effect the "Technological Acceleration Anxiety" which will only increase going forward.
This technological disruption is unique compared to any prior as there is no end. There is no period of stabilization and adaption. It is continuous and accelerating. Human beings need some islands of stability to plan and reason about their lives. This is going to become increasingly difficult.
I've spent a lot of time thought exploring potential societal impacts. I'm not sure how we avoid the negative ones, but I keep listening, exploring and thinking.
Much of my thoughts I've collected here, in case you have any thoughts or feedback.
Mark, for ecommerce entrepreneurs (i.e. non-tech/coders) what should I focus on now to be able to make money with AI? Thanks for any advice.
As I heavy user of Midjourney and ChatGPT for my own use case and for my clients around the world, the new AI landscape is a huge advantage for my workflow. I have not used stock image sites for the last 5 months, being able to create all I need with AI :-)
My side gig is helping environmental and disadvantaged folks manage (formerly hacked) WordPress sites for their causes, which is why I came to know Wordfence, but my day job is as an aerospace engineer, systems engineer, and technologist.
I chuckled at the initial responses to technical questions posed to GPT-4 and, clearly, I would not put my life on the line to fly in anything made by an AI tool...today.
However, at a more granular level, aircraft are just collections of parts flying in formation under mutual agreements and rules set by the designer. All of this can and will be learned by an AI tool.
I'm too old now to change horses far (though I do know a bit of Python), but I am definitely in a position to ask hard questions about how systems are designed and how companies are taking advantage of these concepts in optimizing their approaches. When squeezing the capability out of every drop of energy through proper design matters, we need to take every bit of the opportunity to explore the potential of AI.
Even at 10X per year, I may still not make it to fly in an AI-optimized electric air-taxi, but my casket might.
Thanks for the perspective on technology. The concepts transcend discipline.
Thanks for the excellent article you posted.
Over here in Italy the Privacy protection authority (the 'Garante') vetoed chatGPT for the time being, and right now it's inaccessible from Italy, because of concerns regarding not appropriation of someone else's work (other authorities are looking into it) but because in order to train the tool, someone harvested the net for personal data without asking for permission and without informing whoever owned the data. That in Europe is a crime. But so far only Italian authorities seem concerned.
My concern is that we have a dilemma. Either you decide to stay out of the training process, refusing to give (seems that ChatGPT owners are mindful of these vetoes) your images and data and experience etc to the AI engine to train it, and in that case you risk oblivion in a very short time (like in "what's not online for young people is nonexistent" becomes "what AI doesn't know is nonexistent"). Or you help the training, and in that case when someone asks about you they'll get a correct response... but you'll give up your creator rights in doing so.
I don't know if you ever heard of Lucio Battisti. He was in the 60s-70s the most famous Italian singer, possibly was the greatest all time, millions of records sold, huge fame and so on. He died young, and his wife was obsessed with the authenticity of his work, defended it to the point of forbidding covers of any kind, by anyone, on any medium, even forbid remastering the original tapes. The result is that italians born from the 90s on hardly know him.
Do we risk the same fate with AI fooled (or deliberately trained to get certain results) by limited/targeted training data as we would get with SEO cheating?
Sounds like Google and whether you should let crawlers index your content, but without the clicks.
Appreciate the article. You got my attention!
Former software developer here. I quit the industry and went into writing, editing, and proofreading. Call me a dinosaur, but I refuse to believe that AI writing will equal a human’s writing during my lifetime. There are subtle nuances in the language that I don’t think you can teach a machine. Take sarcasm, for example. Even humans sometimes struggle to perceive sarcasm in writing. So I can’t see this happening in AI anytime soon.
As an author, I have heard about books coming out onto Amazon that have purportedly been written by ChatGPT. It’s disappointing because these people are not interested in providing service or helping customers. They’re simply trying to make a quick buck by churning out poor quality books and slapping them onto Amazon. (I’m sure AZ will crack down on this at some point.)
So far as my own life is concerned, I remain sceptical and I’m pushing back on AI as hard as I can. I won’t speak to a computer or a bot, if I can avoid it. I sometimes get them on customer service lines. I got one several weeks ago that frustrated the living daylights out of me. So I bellowed back expletives at it, and within seconds, I got a real human being to talk to. (I think it was Argos.)
This is the first time that I read an article in my Wordfence emails (they usually get directed to a folder called "read later", but never). I'd just heard about MidJourney but has yet to try it. Thanks for the insight.
You’re very welcome!
It was supposed to be "Frodo" and Sam across The Dead Marshes instead of "Bilbo" and Sam across The Dead Marshes, right :-)
That’s funny when I wrote it I thought Frodo not Bilbo and then wrote Bilbo anyway. Leaving uncorrected for posterity.