Ooo shiny things

In Technology we always used to think in 3-5 year cycles. Mostly because it involved hardware. It was very handy for total costing and giving a timeframe to determine what was next. To think in terms of a lifecycle.

The age of Cloud and SaaS has turned that on its head a bit. Now it’s just a monthly/annual bill and I sometimes wonder if the ‘lifecycle’ consideration has come down to just that point of time at renewal.

Do we look at 3-5 years ahead anymore?

Sometimes it’s all about product, data and use. I was reading about the furore over Adobe updating terms of use for its products which seemingly allows it to use the content you create. For machine learning and to improve its own products…big shock

Adobe users upset to find company has been scanning content

Update that triggered web rage didn’t really change much – Photoshop slinger has always been able to scan your stuff

The RegisterBrandon Vigliarolo

I think we’ll see more of this from organisations. I read this paper ‘Will we run out of data?’ and their estimate was to start getting concerned from 2026.

Will we run out of data? Limits of LLM scaling based on human-generated data

Mascot SammyPablo Villalobos Anson Ho Jaime Sevilla Tamay Besiroglu Lennart Heim Marius Hobbhahn

LLMs need new data. Synthetic data I think as inbreeding - nothing good will come from that. Most of the Internet has been scraped so companies will look for more private data. One to consider over the next 3-5 years if you’re on SaaS or Cloud.

Then I read this article by Benedict Evan’s which gave an interesting breakdown of the problem of how do you bring this product (AI, GenAI, LLMs whatever) to the mass market?

Building AI products — Benedict Evans

How do we build mass-market products that change the world around a technology that gets things ‘wrong’? What does wrong mean, and how is that useful?

Benedict EvansBenedict Evans

No real answers but there are hard questions to solve. The elephant in the room is the product being pushed into solutions it’s not brilliantly designed for. But billions are on the line.

Product, data and use - problems associated with all of them over the next 3-5yrs. Some tough decisions to make, and a lot on the line.

But is it even the most important thing right now if you remove the hype? Does it require this level of attention and funding? Because of a cyber-attack an appeal had to be launched for blood donations

NHS appeals for blood donors after London cyber-attack

The NHS says last week’s cyber-attack means blood cannot be sorted at the same frequency.

BBC NewsJames W Kelly & PA MediaBBC News

and junior doctors are being asked to help out with hand delivering blood tests

London hospitals: Trainee medics urged to help after cyber-attack

Students are asked to act as “floor-runners” to help London hospitals recover from a cyber-attack.

BBC NewsSharon BarbourHealth CorrespondentJess WarrenBBC News

But keep looking at the shiny thing by all means.