AI Brainstorming

Monday is focusing on AI. We'll discuss what topics to consider.
Submit your questions anytime www.askofficehours.global
Join our global community officehours.global
00:00 Intro
10:53 AI has some time saving features but mostly I use it to help me edit writing. I do like using custom bots but how will we get over the LLM hype and get actually useful AI? Is it RAG?
14:25 Battle of the LLMs: Bring on guests from the 3 major LLM makers (OpenAI, Google, Anthropic) to talk about why their LLM works the way it does and their advantages and disadvantages.
16:59 Keeping up with text to video models like Sora and Runway's Gen3, and others. Staying up to speed on how they fit into the workflows.
18:39 Live demos of our favorite time-saving AI uses. Each demo would be under a minute to show the context, the previous process without, and then the process WITH the tool.
19:41 We should have Daniel from Fabric on to talk about his incredible open sourced prompt templates. [link]
21:11 What personality should AI mimic? ChatGPT says it will be in a naturalistic voice like Scarlett Johansson in the 2013 movie "Her" - July 1st NYTimes - [link]
28:36 I've used Claude a lot to aid planning and brainstorming, and one barrier that can affect ones' workflow (with any LLM) is the usage limit. Do you think that we'll see said limits be a thing of the past as CPU/GPU power on the desktop increases?
31:05 What to expect from AI impact over the next year? What workflows and problems are ripe for change sooner vs later?
32:37 Considering that most of the generative AI conversation is about unreleased features, how will panelists have meaningful conversation that is deeper than speculation?
33:59 I'd like some second hours on AI 3D generators (e.g. LUMA Genie, Meshy. etc.).
36:02 How A.I., machine learning, neural networks, are being incorporated into dedicated processors and how that will affect future camera technology and production.
37:42 How about 'Avoiding AI Misuse and Mistakes'? How to guard against you or a teammate relying on AI when they shouldn't how to catch errors and hallucinations?
40:46 is there an A.I. aggregator? A single interface where I can see the responses from several models for the same inquiry?
41:57 Local AI chip sets that are being built into Intel processors and Apple... etc
42:58 What about how the major vendors are shoving AI into their specific products (Ross, Matrox, Sony, etc...).
45:20 How we approach exploring AI features without losing too much productivity time? (scheduled play time? AI courses? daily? weekly?)
50:36 Topic idea: other types of AI that are not LLMs (or at least not chatbots), and how they are useful today in real-word scenarios.
51:52 I would love to have Michael Cioni back to discuss more about "Utility AI" or just discuss Utility AI as it applies to media production...
52:22 Discuss Yip - Transform text into multiple choice, true/false, or fill-in-the-blank quizzes automatically with our AI-powered question generator.
52:59 I'd like to do a second hour on LTX Studio.
53:22 I think a show on how to protect you IP when using AI would be important for this community...
55:36 I remember the story of a session drummer who was concerned about his job in the 1980s and the advice he got was: "Get a LinnDrum. Learn how to program it." Will we see more scenarios like that with AI?
58:24 Discuss LiveReacting AI - Run Interactive Live Streams that can be applied to teaching and gaming.
58:43 AI features that will be included in Office Hours 2030 or What's in J.J.'s brain? ;)