Ok – I could be ‘old man shaking his fist at Cloud’ but i’m really struggling with the AI thing
I have 3 main points to make about AI and all three of them vary on their strength per industry, but realistically the whole buzz is bullshit generally.
I have yet to find a broad appeal use case that has the following traits:
- Its is a relevant use case now, that is also sustainable into the future
- it cannot be replicated by another dev with access to the same toolsets that everyone else has (chat gpt etc)
- ai startups have structured IP that allows them to uniquely commercialise
- have a commercial model that makes ANY sense
Now, in 2017 I was at the Salesforce Conference called ‘Dreamforce’ and it was a cacophony of people running around talking about tech and I’m convinced there would have been only 20% of them that had actually built anything, coded anything or had been in the trenches making tech work. Now having said that Dreamforce put on an amazing set of panels over the 4 days and the type of customer/tech innovator that we got to hear speak was outstanding.
What stayed with me was the panel discussing the future of AI and what it meant for us all. Amazingly the Head of AI for Google (i think it was Hinton at the time) was there and was answering some pretty amazing questions, and on the same panel Will.I.Am and he was absolutely as you would expect as a contributor to an AI panel…like a post turtle…noone knew why he was there, including him, but he was there and the panel began.
Dr Google spoke about the formula for AI which I use to this day which is:
- AI = Data + Machine Learning + Human in the Loop
- See below
To this day I argue till im black and blue in the face that this is a crucial piece when discussing how ‘AI IS THE BIG DISRUPTOR’ or ‘CHAT GPT IS GOING TO REPLACE EVERYONE’ or ‘WHAT HAPPENS WHEN WE ALL GET REPLACE…..ALSO WHY DO I HAVE THIS RASH?’….humans are still a massive component to the operation, curation and the deployment of AI. But then I think about Will.I.AM
Superflous punctuation in his name aside, Will.I.AM (is that right?) is the current model for AI and the current buzz. We arent sure why its big, but we know its a something. The old versions are way truer and much better than the current versions and less flashy industry ‘WHOO HOO’ and certainly a lot less hot air buzz. It has been commercialised in a way that is unsustainable, is more brochure than substance, EVERYONE is talking about it (recent BEP is terrible but marketed hard) and then theres addons, like his headphones that just are stupid and he wouldnt take off on the stage.
Diatribe yes, intolerance maybe, but the worst part about this buzz is the following:
- The same toolsets are being used by everyone – not much innovation is being done in a unique commercially sustainable way
- Everything has a ‘me too’ (not the same) feature set that crops up in weeks to do similar
-
Noone is really sure how to commercialise it properly
-
Content archives cant monetise selling AI tools and scanning for metadata insights
-
Nothing says ‘i the possibilities are endless’ like someone that has a speaking tour deal instead of shareholders
-
The one caveat here is that I have heard of 1 commercial use case which is ‘you’ll have 50c a month AI which helps you find ingredients at the shops. $1 a month AI which will help you draft responses to emails. $10 a month AI which will help you negotiate better rates on your software and $10,000 a month AI which will help you creat a genius investment scheme in shares’
The fact of the matter is that a few things needs to happen and one thing we have to remember – we recognise that we need a human in the loop for AI to be useful to curate data, define our outcomes and what we are looking for AND anyone that is getting ABSOLUTELY RAZZED ABOUT AI doesnt need to prove anything to shareholders but absolutely needs to book time on a speaking tour