The decision to integrate generative AI can be framed as a simple no, a clear yes, or a more nuanced "chicken and egg" problem for developers to solve.
Avoid adding AI for hype by using the desirability, viability, and feasibility framework to ensure it solves a real user problem and has a business case.
Generative AI is a clear yes for industries like customer service, data entry, and translation where it can fundamentally disrupt and improve existing processes.
Developers must provide technical perspective to business stakeholders and investors who are often driven by market hype rather than product need.
Use Retrieval-Augmented Generation (RAG) for querying documents and reserve fine-tuning for teaching a model a specific style or vocabulary.
Hyperscaler platforms like Azure are rapidly simplifying AI integration, moving from complex custom builds to more accessible, business-focused solutions.
The underlying transformer architecture of LLMs is now being applied to new domains like visual data and time series forecasting for finance and IoT.
Generative AI is evolving from a simple tool to a programming assistant and is now moving towards autonomous multi-agent systems that can model entire companies.
Developers must act as partners to business teams, providing the deep technical knowledge needed to make informed decisions about when and how to use AI.
59:39Julian Joseph
25:17Scott Hanselman
29:26Alexander Birke & Silke Eggert
28:38Daniel Tao, Kathrin Schwan, Faris Haddad & Florian Schaudel
26:54Duan Lightfoot
30:24Markus Walker
21:22Kapil Gupta
29:00Hanna Hennig, Michael Ameling & Tobias Regenfuss and Mike Butcher
Jobs that call for the skills explored in this talk.
Conrad Electronic SE
RE-INvent Retail GmbH