š«§ AI Is a Bubble
Good morning, friends!āļø
I am again writing this newsletter after finishing a mixed-weather 13km round Green Park this morning. For those of you who donāt live in London, Green Park is the park by Buckingham Palace, and no, I didnāt see the King.
I forgot to mention last week, but this newsletter recently crossed 2,000 readers!
Thank you to everyone who reads it every Monday, and I hope it helps you in any way, even if you just listen to the podcasts or read any of the books I often mention at the end.
Onwards and upwards! š
Right, this week, I want to have a little rant about AI. I want to preface that I am quite bearish on AI at the moment and think it is overhyped, so naturally, I am quite biased.
Earlier this week, I read the following article in the Guardian:Ā From boom to burst, the AI bubble is only heading in one directionĀ byĀ John Naughton. I have to say, I pretty much agree with everything said in the post.
No new technology has sustained exponential growth forever. The classic example is theĀ dot-com bubbleĀ at the turn of the century, but bubbles pre-date that.
There was theĀ railway bubbleĀ in the 1840s,Ā the radio bubbleĀ in the 1920s, and theĀ electronics and transistor bubbleĀ in the 1960s.
Perhaps the most recent example is cryptocurrencies, NFTs, and blockchain. They were all the rage a couple of years ago, but now you barely hear a peep about them.
Let me be clear, the technology that sparked these bubbles is undeniably valuable. Railways revolutionized transportation, radios enhanced communication, and transistors are the backbone of modern computing. I still see potential in crypto, I even invest in it, and believe blockchain is a solid technological concept.
The problem is always hype and overexaggerating what the technology can do, which is what I think is happening with the current āAI boom.ā
As a data scientist, I donāt like using the term āAIā as it normally refers to any decision-making algorithm, such as Googleās search engineĀ page-rank algorithm, theĀ recommendation algorithmĀ on YouTube, or even simple email filtering.
The point is that AI has been around for a long time. This current AI boom is around Generative AI, where the technology isĀ generativeĀ instead ofĀ discriminative. I will spare the mathematical details, but it's a different statistical process where the generative models generate a range of outputs.
I think the current GenAI models are amazing, and I am overly surprised by how good they actually are. Just look atĀ OpenAIās recent GPT-4o; it isĀ truly special.
The problem lies in the massive spike in āAIā companies now, which is basically fuelling the bubble. Just by putting āAIā in their name, companies get attention and catch the wave, exactly what happened with the dot-com boom.
Like the Guardian article says at the end:
āSo, going back to that original question: are we caught in an AI bubble? Is the pope a Catholic?ā
Weekly Favourites ā¤ļø
š©Ā Newsletter ā Michaelās Newsletter.Ā I have been reading quite a bit of
ās work and it's very inspiring. He was a software engineer atĀ Netflix, earning $450k, and left to pursue his own ventures. His newsletter covers all his learnings from engineering, entrepreneurship, and business.āļøĀ Article āĀ From boom to burst, the AI bubble is only heading in one direction.Ā Considering this article inspired this edition, it is definitely a favourite this week.
š¬Ā YouTube āĀ Luke Made It.Ā I love chill and relaxed day in the life videos, and Lukeās hit the spot. I have been following him for a while and absolutely love his content. He is also an engineer but clearly has fantastic videography skills!
(PS: Some links are affiliate links that I get a kickback from with no extra cost to you š)
My Latest Content š¬
You can reach me on:
LinkedIn, X (Twitter), orĀ Instagram.
MyĀ YouTube Channel and MediumĀ Blog to learn technical data science and machine learning concepts!
š” If you are interested in sponsoring this newsletter see here.