What’s the biggest gap you see right now in generative AI tools? What type of tool do you wish was on the market?
CISO in Finance (non-banking), 10,001+ employees
Having inbuilt adequate Security controls which can protect the information from misuse and breach . Have not seen much information or commitment from vendors in securing this part. From the output perspective they give generic results for most of the queries and it does not give the exact results which business teams need. Lot of information need to be fed in for tool to understand the requirement and before it can be use for production. Integration with in-house developed applications I see a challenge.Director of Information Security in Telecommunication, 10,001+ employees
Generative AI models, like GPT-4, have shown impressive capabilities but still face challenges. They often lack a deep contextual understanding, producing outputs that may be coherent but lack nuance. Additionally, while customization is possible, it requires significant expertise, and the models can sometimes reflect biases present in their training data.On the wishlist for the market is a tool that seamlessly combines the flexibility of broad generative models with real-time user collaboration. Such a tool would allow the AI to generate content, seek user feedback, adjust its outputs accordingly, and be domain-aware, ensuring expertise in specialised areas while also being transparent about its operations.
Director of Technology Strategy in Services (non-Government), 2 - 10 employees
It's a combination of prompt engineering and trustworthy outputs.Prompt engineering is a skill all employees should be building, but there is an appetite to push it towards a dedicated role, which is a very short sighted as it will lead to bias (both concious and unconcious)
Trustworthy outputs falls at the other end of the spectrum and is simply that we can't take anything that Generative AI delivers at face value. It needs to be valdiated and checked.
CDO in Software, 10,001+ employees
Tools to agnostically evaluate, privacy, transparency, bias and model energy consumption. A sort of Nielsen rating for retail but for LLMs.Content you might like
Director of IT in Education, 10,001+ employees
Learning, Pseudocode, Code completion, quick answersScalable AI49%
Composable Data and Analytics40%
XOps22%
Data Fabric35%
Engineering Decision Intelligence25%
Augmented Consumers6%
Edge Computing25%
220 PARTICIPANTS
That said, I do think there is a place for it to help in the SOAR front and for finding connections in threats that a human wouldn't see. In addition, it should also be able to help find ways to remediate and clean up inefficient implementations of products like firewalls, etc.