Скачать или смотреть Spencer Kelly's AI Masterclass - this is what happened

I’m really proud of Spencer Kelly’s AI Masterclass with Michael Pound - we discussed in-depth all of the latest topics and issues...

Practical demos including:
➡️ Use AI in a smart way to manage your inbox.
➡️ Properly control AI image generation

What are organisations doing wrong when implementing AI:
➡️ Going in blind, throwing AI at a random problem
➡️ Instead, come up with small concrete examples of how LLMs can help, and test them in a controlled environment first

AI’s unfixable, fatal security flaw:
➡️ LLMs don’t distinguish between instructions and data
➡️ This means there’s no way of stopping someone sneaking in commands that can persuade an LLM to go rogue

How should coders work with AI:
➡️ Think of LLMs as pretty good, but inexperienced, coders
➡️ They can only make small tools and utilities on their own
➡️ The better coder you are, the better you'll work with AI

Will AI remain “free”:
➡️ Is this a standard disruption model, a land-grab of users, with charges brought in after we’re locked into an ecosystem?

Is there something different about Grok that allowed it to create inappropriate images:
➡️ All image generators are capable of creating those images
➡️ But most have better guard-rails at the front end

I’ll drop clips from the session in the coming days. And if you came to the Masterclass or you’re a newsletter subscriber, you’ll get first access to the whole recording, plus an extra Q&A.

I talk at loads of events about AI and everything else tech, I interview execs, and also train them on how to do what I do on camera and on-stage. If you’d like to know more, get in touch!