Why You (and Your Company) Need to Experiment with ChatGPT Now

Tue Mar 14 2023

Chat GPT AI

Chat GPT AI is a tipping point for AI and could increase productivity by 50%.
Companies need to develop strategies, policies, and best practices to harness the power of Chat GPT AI.
The technology is not new, but the size of the models has made them much more useful suddenly.
Preliminary studies show 30-50% improvements in productivity for people not trained on the system.
Applications include writing computer code and documents, generating marketing ideas, and performing Porter's Five Force analysis.
Individuals can get significant performance differences with Chat GPT AI today in their work.
There are no rules or manuals to work from; companies have to explore opportunities and options themselves.

There are secret users of chat GPT in every company who have cut their work by two-thirds.
People spend about an hour to work with these systems, and I've talked to tons of people who have automated things like grant applications or regulatory compliance.
Chat GPT is a general-purpose technology that cuts the most boring parts of your work.
Companies need to realize that there isn't somebody who knows anything, but at the same time, this technology is happening so quickly, and the performance improvements are so large.
Companies should assess their exposure to this and understand where it can benefit them.
Companies with strong innovation frameworks are better set up to get rolling with chat GPT.
Organizations should be in triage mode and use a war room approach because the situation is evolving so quickly.
The policy will come from the use case. In cases where there's high levels of accuracy required, then current versions of chat may not be what you need. But specialized disruption is coming next.
A marketing consultant company could use chat GPT for market research, writing articles for SEO optimization, and creative services.

Using chat GBT for creativity and innovation
SEO article writing can be done using chatbots, but human oversight is required
AI can be used for research in marketing consulting
Hybrid work with humans interacting with technology can lead to more productivity and better work
Chadji BT can make everyone more productive or level the playing field for worst performers depending on the organization's setting
Managers should experiment with their jobs by automating tedious tasks through chatbots to see what works best for them
There is no need to be intimidated by AI as it is a tool that can be programmed in English language prose

There is no advantage to being a coder or understanding how large language models work when using this tool.
These tools are programmed in English language prose and can automate tasks like writing market research reports, letters of recommendation, and marketing slogans.
New technologies tend to first disrupt jobs and then add jobs, so workers should figure out how much of their own job they can automate to understand their exposure and risk to these technologies.
The first things people do with this tool often give the worst possible answers, but as users explore more, they realize it can automate huge amounts of writing and other tasks.
The risks associated with these tools include hallucination, misinformation, chatbots that are very convincing, automation of people's tasks without clear benefits for workers or companies, and larger implications for machines that can perform intelligent tasks without sentience.
Other companies are working on different applications using natural language processing and large language models. OpenAI is releasing GPT-4 soon which will be a game changer. Google has the largest research team on AI and is cautiously rolling out products specialized for different industries.

There are many LLM's and large language models available, including diffusion-based models for creating images.
Tools are being developed to create videos and user interfaces based on text descriptions.
Specialized tools are being created for meetings and writing assistance.
Organizations should start experimenting with these technologies while complying with ethical and legal obligations.
Dangerous or illegal use cases should be avoided.