Related Posts
What's a guilty pleasure you indulge in at work?
Anyone work or has worked at MGO?
Is Google the end game?
New to Fishbowl?
Download the Fishbowl app to
unlock all discussions on Fishbowl.
unlock all discussions on Fishbowl.
What's a guilty pleasure you indulge in at work?
Anyone work or has worked at MGO?
Is Google the end game?
This site is protected by reCAPTCHA and the Google Privacy Policy and Terms of Service apply.
Download the Fishbowl app to unlock all discussions on Fishbowl.
Copy and paste embed code on your site

Scan your QR code to download
Fishbowl app on your mobile

Riding this out for another 5 years before the whole career path and industry goes up in flames
Retiring.
Get a MacBook Pro with as much memory as they offer. Install LM Studio, Ollama, and Python. Sign up for Hugging Face. Download models to run locally. Play around with prompting. No code required. If you want to build an "Agent" it's like 10 lines of code - just a loop, some tool/interface definition, and a line or two to keep glomming stuff onto the prompt.
I vibe coded an app that searches through my PowerPoint decks and PDFs. So when I want to find AI use cases specific to an industry, an approach slide or a specific architecture, I just type my question and the relevant slides show up. I did this locally on my MacBook Pro. Fantastic learning experience.
I switched from consulting to hands-on build an AI company. The pace of change will be so rapid that you need to be part of the change to truly stay relevant. At least that’s my bet…
I was part of publishing AI reports at my former company and saw first hand how little we knew in practice :)
Building an agent to take a first pass at some regular slide content.
Coach
Just following our experts on the USMC.
Reading, podcasts, building
Experiment, especially with tools that can actually be useful for your line of business. Forget the ”prompt engineering” trend and start thinking actual workflows like presentation making (try Gamma, but a bit hit and miss), interview analysis (try Skimle, but still requires manual review) and financial analysis (try Claude’s new tools, work well).
Then on top of personal AI use (table stakes), get smart on real life examples of use cases relevant for your clients and how they work in practice. Reading helps here, especially from more tech oriented people hanging out on their own forums as the glossy case studies are just bullshit.