Intelligence (AI) and Machine Learning (ML) are rapidly transforming how software is designed, developed, and tested. With assistants like GitHub’s Copilot helping developers feel more productive and faster; everything from code creation, testing, and bug-fixing are being disrupted. What does a modern software development team look like?
In this podcast episode, Denys Linkov, a Machine Learning Team Lead at Voiceflow, discussed his journey to becoming a team lead in AI and ML. He also talked about the challenges faced by AI platforms and how Voiceflow handles them, including their approach to implementing Chat GPT and their stack, which includes a real-time collaborative canvas, a custom programming language, and an ML side that uses both third-party and proprietary models.
Denys also discussed the impact of low code and no code tools on the demand for developers and the need for continuous learning and upskilling to stay relevant in a fast-changing world. He emphasized the accessibility of AI technology for startups through open-source models and closed-source APIs.
Moreover, Denys shared his concerns about the potential social harms of AI and ML and the need to modernize systems before implementing these technologies. He advised aspiring developers to build things in practice, use the tools they are interested in, find mentors, and show ownership of the interview process by bringing a portfolio of digital and printed work.
Finally, Denys discussed two techniques for improving AI's focus on specific data and mentioned a proof of concept bot built on Voiceflow that retrieves information from various sources and summarizes it using a large language model.
Index
=====
00:00 Intro
00:48 Journey into AI & ML
02:03 Denys Switch from programming to Leading Teams
03:18 Does he still code?
05:15 Poll--The Biggest Challenge with AI Platforms
06:02 Rate the importance of Challenges
09:56 How Voiceflow Started using GPT
13:43 The Stack at an ML Company
16:27 What was difficult or easy about implementation?
19:47 Transition from GPT-3 to GPT-4
22:45 How do you deal with Hallucination and negative effects?
24:28 How it's affected the development team
29:30 Impact of Low Code and No Code Tool in Developer marketplace
35:27 Historical Examples of Technology Changing Job Marketplace
41:42 What Should Companies do to implement AI responsibly?
44:41 What's important for those getting into the field.
54:05 Advice for Interviews and Internships
01:01:59 Handling More Specific Models
Links
=====
Voiceflow: https://go.raybo.org/TFIT-1201
GitHub Copilot X: https://go.raybo.org/TFIT-1202
What will the AI stack look like: https://go.raybo.org/TFIT-1203
Denys Linkov Recommends Students Build APIs: https://go.raybo.org/TFIT-1204
Chat GPT Outperforms Crowd-Workers PDF: https://go.raybo.org/TFIT-1205
Denys--Disgress with Collapsing Costs of Software: https://go.raybo.org/TFIT-1206
Denys--Will Chat GPT Disrupt Industries: https://go.raybo.org/TFIT-1207
/ denyslinkov_can-gpt-4-disrupt-industries-y...
GitHub Copilot for Business Claims: https://go.raybo.org/TFIT-1208
Book--BS Jobs--A Theory: https://go.raybo.org/TFIT-1209
Hugging Face: https://go.raybo.org/TFIT-1210
Смотрите видео How Machine Learning & Artificial Intelligence Change Development онлайн без регистрации, длительностью часов минут секунд в хорошем качестве. Это видео добавил пользователь Ray Villalobos 01 Январь 1970, не забудьте поделиться им ссылкой с друзьями и знакомыми, на нашем сайте его посмотрели 23 раз и оно понравилось людям.