The cost of intelligence is diminishing towards zero.

Today we’re launching Sido Research, a nonprofit AI research and deployment organization built on two pillars:

1. Accelerate safe AI adoption across Hong Kong’s local small-to-medium enterprises (SMEs) and family businesses, pro bono.

   
   

2. Make sense of Hong Kong’s AI future from a strategy and governance lens

We stand at the cusp of a cognitive revolution, where artificial intelligence will become a general purpose technology much like electricity and the internet. Hong Kong must be well equipped.

• We should not compete on the LLM frontier, but primarily on adoption strategy and economic benefit

• We have a unique opportunity as an expert data and reinforcement learning environment marketplace given our service-based economy (93.5% of GDP)

• We need proper governance and a strategy for compute infrastructure + energy for inference

“Andrew Ng + OpenAI + a16z partners leaving and starting AI services / advisory firms... huh, lesson there probably.” ~ @EvanLyseng

SMEs and adoption is the biggest opportunity of our time

We’re naming ourselves Sido - a nod and romanization of 士多, the local corner store that powers HK’s economy.

Small-to-Medium Enterprises (SMEs) in Hong Kong account for 98% of all enterprises (360,000+), 50% of the GDP, and employ 44% of the labor force. These are the guys Accenture and McKinsey won’t touch. They aren’t trying to expand into new revenue streams or capture market share, but nonetheless should not be left in the dust for this revolution.

We’re targeting the old industries and niche verticals.

Taking Palantir’s forward deployed engineering approach, we are working, strategizing and building software alongside businesses. Our conviction is that the bulk of AI’s benefit will come in the deployed application layer. Continually, the HK government recognises the importance of SMEs and we see an opening for a team of scrappy forward-deployed engineers to swoop in and drive efficiency and economic growth.

We believe AI adoption can take place on two fronts:

1. On their revenue driving business: AI is increasing headcount without extra cost. There will be better margins.

2. On their backend operations and logistics: to the local Grandma who runs her bakery, automating her logistics agentically means more time for her to spend time with her grandchildren.

Messy data. Broken pipelines. Manual labour. Figuring out what model to use. Integration is easier said than done, and so we’re running an initial trial with 5 SMEs, writing a case study for each as we go.

To VC and PE investors, we’re simultaneously using this as an opportunity to understand the paradigm of AI-driven acquisitions and rollups in HK, especially given the aging population. We're seeing Chamath's 8090.ai, Thrive Capital's Thrive Holding and General Catalyst's Creation Fund - there might be interesting arbitrage opportunities here in HK.

Hong Kong is yet to have a coherent strategy

Despite buzz and talk, we’ve yet to see a roadmap and well defined plan for HK when it comes to AI. We’re seeing increased directives from the Policy Address as well as questioning in the Legislative Council - as a result we aim to dedicate research effort on understanding governance, providing opinions and advice to policymakers and legislators.

Equal playing field

Given AI's rapid and novel evolution, the role of a policy advisor here is distinct. Expertise isn't confined to traditional policy experts; a 20-year-old can grasp it as readily as a tenured professor.

The US is competing on the frontier. China on the adoption. But with every country and region, having sufficient inference capacity and a suitable infrastructure is of utmost importance.

From establishing an unified strategy to mobilising labour market discussions, we hope to kick off AI discourse.

This is just the beginning. Interested in joining or funding us? Shoot us an email sido@sidoai.org