How are you using AI to simplify your work?

Being a part-time writer AI writing tools have really helped me cut off extensive research time and now I am able to complete 3 hours of work in 1. Plus, being a language enthusiast, ChatGPT has really been helpful to create learning plans. What is it like for you? What and how are you using AI to simplify your work/study?

Hi @mansibhavsar

I would say Github Copilot is the AI tool that I use most often. It’s not perfect, but it helps :slight_smile: (I haven’t tried alternatives (like CodeWhisperer, Google code completion or others for that matter) so I can’t compare, but code completion is one of the areas that I find AI helpful).

Of course there are many “hidden” “AI” that we don’t necessary think we use like product suggestions etc., but if we are talking about stand alone tools then I would vote for Github Copilot :slight_smile:

1 Like

I don’t use any chat-bot tools, because I don’t trust their results.

2 Likes

Hi @TMosh

Since it’s Friday … :slight_smile:

2 Likes

In addition: a great technical product is of course the Google Knowledge Graph: Google Knowledge Graph - Wikipedia, containing ca. 800 billion facts!

It is probably also one of the best business models in the world with an incredible cross margin:


Feel free to check out this article from 2012:

The Knowledge Graph also helps us understand the relationships between things. Marie Curie is a person in the Knowledge Graph, and she had two children, one of whom also won a Nobel Prize, as well as a husband, Pierre Curie, who claimed a third Nobel Prize for the family. All of these are linked in our graph. It’s not just a catalog of objects; it also models all these inter-relationships. It’s the intelligence between these different entities that’s the key.

Source

But of course, GPT4-powered Bing isn’t that bad either! :wink:

Personally, I think the combination of LLMs with semantic technologies is going to be highly relevant in next months and years.

Best regards
Christian

1 Like

well in that case I don’t trust them either because sometimes the results are biased and are based on stereotypes but, the writing work that I do does not involve any research work. So, i just input a few bullet points and the bot returns me an elaborated text.

1 Like

this seems amazing! I’ll give it a try.

1 Like

Hi Christian ! Sorry if this is self evident, I am a total beginner. Are you up to share more on why you think this particular combo would be relevant, for example will it address gaps in accuracy/address lack of trust in users, improve factualness etc? I’m not from a technical background (Philosophy background in AI) so appreciate any context you can give here, thank you!

Welcome to the community, @Tinayazdi!
Thanks for your question.

Have you played around a bit with LLM-powered applications like Bing?

E.g. at Bing the „creativeness“ of the AI or let’s say their „risk to hallucinate“ can be adjusted manually, when you use the bing search.

What I believe is that when (semantically described) facts (e.g. which newspaper, which author, date, etc.) are well linked to relevant context so that also the user can look sources up and quickly validate the answer. So yes, I could imagine that in particular trust and interpretability can be improved as well as the hallucination risk might be mitigated eventually.

Best regards
Christian

Here some update on this note which might be interesting for some of our fellow learners:

Sources:

Best regards
Christian

linking a paper by Deloitte on Responsible Enterprise Decisions with Knowledge-enriched Generative AI -
Why is it essential for enterprise-level generative AI to incorporate knowledge graphs?
which discusses the concepts of combining GenAI and semantic technologies and outlines how a tech stack based on knowledge graphs, leveraging LLMs can help to reduce hallucinations by providing more relevant context and enhance transparency.

I found it quite interesting!

Best regards
Christian

AI can be used in different ways to simplify your day-to-day work. However the results provided by AI tools cannot be totally relied upon as they might not be accurate but we can get some sort of clarity.