Think is an ability gifted to human beings. AI can only perform tasks that are modelled/synthesised by the progress of society such as art, writing, coding, music.

We analyze exposure by industry and discover that information processing industries (4-digit NAICS) exhibit high exposure, while manufacturing, agriculture, and mining demonstrate lower exposure

AI needs human in the loop, thus probably create a new occupation like AI auditor, AI reviewer (in terms of preference and safe behaviours of the model).

Regulations will be in place to make sure AI systems are aligned with human values. Policy makers will need time to understand to models, but my prediction is there will be regulations.

AI systems will need a throughout process of reviews through auditor, human feedback, restriction on data collection, etc.

OpenAI suggestions

Data collection restriction, if being implemented poorly, can harm the development of legit researcher in AI alignment

Trace back for content to identify if content was generated by AI. AI companies and and platforms collaborate together. (AI company identifies AI generated content, platforms flag account).

Eventually, GPT-4 and plugins will be available for public use. What implication does it have?

Instead of searching through Google, use only one universal search that can do anything. What does it mean by anything?

• Chat back and forth to book flight ticket, plan for a trip, looking for cheapest price etc [Personal]
• Connect to other platforms to get insight. Think connect to Github, Discord to gather information for dev portal. Who does this? Maybe Github, Discord to develop plugins?

Disadvantages of Microsoft: It’s too big that it can’t do everything. At best, the Bing search can be a general purpose assistant

How about with the help of plugins? General purpose assistant with the help of plugins can become specialised in a task provided in the plugin. So maybe the future of internal tools will be in form of plugins?

So no code/low code builder for plugins that can bring across different AI models and integrate with other tools. Hmm, but this means plugins need to be published on Microsoft Plugins Marketplace, or ChatGPT marketplace, which is not private to business.

Monetary policy

Main purpose:

• Stability of the currency in Australia: low and stable inflation rate
• Full employment
• Welfare:

Inflation rate in Australia should be at 2 - 3% on average and it has a medium-term flexible rate. That is to keep the in inflation low and allow flexibility for the Reverse Bank Board to make monetary policy decision.

When the economy grows too slowly because of weak demand, the Reserve Bank can loosen monetary policy, such as by lowering the cash rate to stimulate economic growth and employment. On the other hand, when the economy grows too quickly because of excessively strong demand, the Reserve Bank can tighten monetary policy, such as by raising the cash rate to dampen economic activity and contain inflation.

Why raising the cash rate?

“A cash rate is the interest rate that a central bank – such as the Reserve Bank of Australia or Federal reserve – will charge commercial banks for loans. The cash rate is also known as the bank rate or the base interest rate.”

While commercial banks are free to set their own interest rates for borrowing, the rates that they charge on loans and offer on savings tend to be derived from the cash rate.

This means that central banks can use cash rates to encourage or discourage consumer spending, depending on the state of the economy.