AI (Artificial Intelligence) - Anyone Benefitting from it?

[Alex Snyder]

My gurus are telling me the next release wants simpler prompts with goals and guardrails and to give the LLM freedom to guide itself to the goal you stated.

That intel is gold and makes tons of sense! Sounds like the natural evolution of AI.

I look forward to a vocal shift in the flow. If you use Claude Code today, you should be familiar with the Plan to Code flow. What would be super useful for people like you and me, Joe, would be to ideate via voice with an AI. Just talk to it. That would be the "plan" phase. Then a design phase where it outputs something like a mock or storyboard to approve. And finally, a build phase where it does the actual work.

With a flow like that, I can see multiple concurrent projects happening. Way beyond what some of us are doing today.

"Yesterday, AI made me efficient. Today, I'm busier than ever."

 

[Auntess]

I have a BS detector for evaluating companies that talk about AI. If you use this approach it will help automatically sort out some of the worst offenders. Steal it, use it:

If the company's name has AI in it - 1 point
If the domain's url extension is either .ai or .io - 1 point
If they use the words "proprietary", "invented", "patent pending" anywhere on the home page - 1 point
If on their LinkedIn page the % of non technical people to technical people is over 50% (this gets harder to measure with big companies) - 1 point

Anything > 2 points is dangerous territory and you should protect yourself at all times. I am not saying every company that meets this criteria is bad btw. Some are just misguided in the short term and trying to standout. However, the concentration of bad actors is higher than average...





I think the same logic applies when looking at AI tools as well. The branding can sometimes be louder than the actual value. For example, with consumer AI products, a simple test is whether the tool solves a specific problem well or just adds "AI" to the name. A focused product like an ai headshot app has a clear purpose and is easy to evaluate based on the quality of the output rather than marketing claims.

Interesting framework. I think the main idea is useful: a lot of companies use AI as a marketing label rather than as a meaningful part of the product. Having a few quick checks before getting excited about a pitch can save a lot of time.

That said, I’d probably treat these as warning signs rather than a scoring system. Plenty of legitimate AI companies have .ai domains, use words like “proprietary,” or have a large non-technical team because they’re focused on sales, support, or a specific industry. Those things alone don’t necessarily say much about the underlying technology.

The more reliable checks for me are things like: can they clearly explain what problem the AI solves, what data it uses, how performance is measured, and what happens when the model makes mistakes? A company that can demonstrate real customer results and technical details usually stands out pretty quickly.