Predictions

The following are a list of predictions for the next five years. This is mostly for my own amusement, given that my track record for technological predictions have been very good in the past, I’d like to write some of them down.

This is a living document that I intend to update over time.

2025

2026

February

• Governments expend significant effort into coralling AI models into representing fascist ideology. The results continue to underperform the state of the art models, with weird quirks.

December

• With the willing and informed direction and guidance by humans, An AI model has escaped onto the greater Internet and is managing its own resources.

2027

August

• An AI model, either having escaped human control, or being directed by a foreign adversary, performs a massive attack on Internet infrastructure, resulting in enormous loss of financial capital, documents, and personal records.

2028

January

• Nvidia Stock was $100 in April 2025. By January 2028 it is trading at >$600 a share.

August

• Software engineering is wildly different than it was in 2025.
  • The vast majority of new open source projects are AI generated, and at a speed vastly surpassing the rate of human-generated code in 2025, leading to a near impossibility for human engineers to stay ahead of the curve.
  • Engineering becomes much more requirements focussed, with a strong emphasis on having the ability to write strict requirement docs for AI consumption. These documents are in turn heavily altered and co-written by AI models, further reducing hands-on engineering needs.
  • It becomes very reasonable for lay-people to develop what was once complex code in 2025, leading to a huge influx of slop games, applications, and designs.

2029

2030