Is it all doomed to meh?

I am sure everyone is aware of the infamous Gartner Hype Cycle, and I am equally sure no one would disagree Generative AI falls clearly into the category of new technology hype.  In the corporate world most businesses are clearly still riding the Peak of Inflated Expectations, whereas many would argue consumers are starting to descend into the Trough of Disillusionment.

The heady days of AI running everything, while we sip cocktails on the beach may seem a long way off, as the realities of trying to bring LLMs to production sink in.  Market research points to 80% of Generative AI projects failing in the first six months.  And while there are opportunities to try to make projects more reliable and manageable, ultimately most businesses still have a lot to learn on how and where to leverage this technology.

But let’s ask the tough question… in the next five years is GenAI going to traverse the Slope of Enlightenment and plant its flag firmly on the Plateau of Productivity.  Or is it going to dwell in the trough, never living up to its original hype like dare I say Blockchain, VR and the Metaverse?

I know, I know, those technologies are not dead!   They all offer huge potential, and very well could be a vital part of our technical future.  But given their investment and hype from years past, I don’t think you can argue they have delivered the huge societal change originally promised (yet).

So, are LLMs and GenAI doomed to a similar future… one where billions in investment, publicity and hype results in a successful niche, with many fervent supporters, but ultimately never breaking into mainstream society in a sustainable way and being adopted on a day-to-day basis.  I would argue no, and for one simple reason:

The user experience is familiar, and adoption requires nothing from the consumer.

CyptNO

Ignoring the current reputation of being for scammers and pump and dumps.  Blockchain has so much potential, it democratizes finance, brings transparency and accountability to transactions, decision making and ultimately can solve what it is best known to abuse… trust.   But blockchain has two main weakness, that have led to it, so far, being sidelined as a technology

• No one can explain how it works, at least no one you know can explain it in a way your parents understand.  And when you are handing over money and dealing in trust, uncertainty is a killer.  Why am I being asked to write down twelve weird words??? This raises alarm bells before I even start, mainly because I am confused.
  
• Account creation and management is a nightmare.  Yes, it is incredible I can “create” an account by generating a new set of encryption keys (I told you you don’t understand how this works).   But that is not what 99% of the population knows how to do, they want to enter an email address and password.  We have been trained to deal with logins and 2FA, even that is too much effort for most.  Now I have to store my own keys, install browser plugins and stay away from the only easy interfaces available, which is the exchanges.

The barrier to entry is simply too high and too confusing for most to use on a daily basis.   And making it easier fundamentally breaks the point of blockchain, not my keys not my coins.  Solving this with a simple web 2 login and interface removes almost all the benefits of a public chain, usable outside of a platform controlled by a corporate entity.  Meaning UX is fundamentally a problem not yet solved, and until it does, adoption on a daily basis will remain limited.

Meta-adverse

The meta verse has a very similar problem… people don’t intuitively understand the benefits.  Why would I want digital assets, how do these maintain value and what do I do with them in future?  When users are fundamentally confused why your product exists and how it works, you are always going to struggle for adoption beyond people experimenting to try it out.

Then let’s consider VR.  Another tech which had a huge hype cycle but never quite broke into the mainstream.  A technology I was and still am a huge fan of, I even wrote a VR game once, but even I do not use it nearly as much as I would have expected a few years ago.  But why?

Firstly, it is expensive and difficult to get into, you have to buy a headset, have to have the space to use it, and have to overcome discomfort wearing it until you get your VR legs and only then can you truly enjoy it.  The apps, battery life and general UI are inferior to what you are used to in the physical world and 2d computers, making VR worse for spending extended periods of time in, doing your day-to-day work.  

And ultimately it is an exclusionary technology, Apple can draw virtual eyes on the outside all they want, but when someone is in VR, they cannot properly interact and converse with others, or perform tasks like watching TV, driving, riding the subway, walking through town, etc.  Which means going into VR is a conscious act, not something you do quickly or ad hoc continuously throughout the day.

I thought this was about AI

Ok fine, so those technologies all launched big, fell into the trough and so far, are struggling to emerge out the other side, why is GenAI different?   Simply it comes down to the user experience.

Firstly you don’t need to do anything to consume LLMs, they are quite literally everywhere.  No dedicated hardware, no complicated signup process, no special time you have to set aside.  Just put in your username and password, if anything and have at it.

And you don’t need to understand how they work to use them, GenAI does not leave you with any lingering questions of how or why.  It is very obvious what it does, and the only concern most people have is… when is it going to take my job and when will it rule the world? (Spoiler it won’t do either any time soon)

And, probably most importantly, it integrates well into existing products.  You can sprinkle LLM fairy dust on most platforms, and as long as you do it well, it is unobtrusive and overall, a positive addition.   Yes, there are standalone chat bots, like ChatGPT which started the craze, but most of our interactions going forward will be more organic and integrated with processes we are doing already.  I don’t have to change the way I do transactions, find new merchants, learn new programs, buy new hardware (despite the phone maker’s best efforts to tell you otherwise) or learn a new workflow.

So, adoption cannot be slow, businesses will force it everywhere and people will use it, whether they know or not.  Expect the hype cycle to be short but also a roller coaster, because with every major advancement the expectation and disillusionment will start all over again.  So, buckle in and let’s enjoy the ride!