Attention: As of January 2024, We have moved to counting-stuff.com. Subscribe there, not here on Substack, if you want to receive weekly posts.
As with all paid subscriber posts, this is me thinking things on my mind through writingā¦
It was bound to happen ever since the AI image generation tools and the recent explosion of Large Language Model powered tools sprung into the public awareness in the past couple of years. ChatGPTās explosion in popularity and the headlines it spawned just put us over the tipping point. The term āAI warā and other ridiculous battlefield imagery has taken hold of the imaginations of reporters even in the mainstream non-tech media. The hype train has hit critical mass and weāre all trapped on the resulting ride.
The media clearly loves war imagery, because running queries for ā$Term warā seems to indicate that (aside from the actual bullets-flying real war in Ukraine) thereās an EV price war going on, a tech/microchip war between the US and China, and various PR wars between people and entities.
Note: While my employer is one of the companies at the center of this narrative, I donāt have any insight over what the heck the myriad teams, executives, leaders, are going to do. Plus, as a relatively late adopter of new tech (I donāt have energy to deal with early bugs) I donāt really have some grand vision of some AI-powered world like endless influencers (gag) and tech writers seem to have been spamming us with already the past few months.
Instead, I just want to reflect a bit on how the next couple of weeks, months, years(?!) are going to be a big annoyance to the data community as the media reports on all this with baited breath and the hype cycle marches towards the inevitable fevered peak.
It was already pretty bad for many of us who work tangentially with machine learning things to see the AI hype ratchet up in distinct steps since even before AlphaGo surprised people by beating Lee Sedol in 2016. Machines got to flex how they were better than humans at a once insurmountable task ā but it still took an expert team to build a system to do it. But now AIās become exponentially more hyped since because itās finally reached the hands of average non-technical users and the output is leaking into public discourse.
As data scientists, weāve already experienced first-hand the relatively quiet but constant competition within the AI space for much longer than the general public. From the big fuss about recommendation systems around when Netflix and Spotify rose to prominence, to seeing AI advances make machine translation significantly better, to seeing how AI/ML powered features very quietly slipped themselves into things like photo/video editing software. Weāve thrown ourselves in the center of the thorny ethics surrounding AI, stood horrified at the quality and source of data used to train such models, struggled with the tech and infrastructure enough to create new job titles around them, and cracked jokes about our regression models getting the AI branding for VC funding.
Keep reading with a 7-day free trial
Subscribe to Counting Stuff to keep reading this post and get 7 days of free access to the full post archives.