@atmoio Dude, I followed your unethical guide to surviving AI and it is working just as planned! I am now leading an “AI Office Hours” for all of Operations at a major US company. I was added to an AI academy by the CTO. Can’t believe how on point you were… 😅
@paulsaladinomd collagen, which is useful structurally but weaker for muscle protein quality.
So on protein density, sugar, ingredient simplicity, and complete-protein quality, Chief looks like the cleaner macro win.
Not a dessert bar, but definitely a better protein bar.
@paulsaladinomd no preservatives, and a short real-food ingredient list: organic grass-fed beef, amaranth, currants, coconut aminos, vinegar, salt, spices, lactic acid.
Most importantly, Chief’s protein is actual beef: a complete protein with all essential amino acids. Lineage’s 20g includes
Interesting idea, more specialized/smaller models focused on one piece of the task. If setup correctly upfront it would make targeted fixes much easier. Knowing upper level management they’d want to save money by using multiple 3rd parties whose agents don’t play well together for proprietary reasons. I’m going to ponder how that would work because I believe you are correct. Create a baton race of sorts. Even though one big model can build it, the segmentation would make fixes far easier IMO. Thanks for the idea!
Hey man, you have a real talent and I deeply relate to your insights. Thanks for making your videos!
My experience has been this at a big financial org - I build something faster and more ambitious, they like it and want more. The more I build, the more I have to maintain. Instead of being intimately tied to a few projects I’m burning out on fixing a bunch of ai based projects I’m not truly connected to.
Also, because regulations and requirements change so much I end up erasing all my initial build time savings gains by having to reprompt for trivial items that break in the process.
Morale of the story - AI doesn’t take away any work it just enables you to take on a backlog that was waiting for you and all of that backlog suddenly in flight means many more problems to fix later on. C suite says “oh, you built that neat app for them? Now I’d like you to build one for them, them , them and them.” The can of worms got way bigger.
Also, I can build out elaborate automations that take a day or two and leverage AI but vendors change, procedures, workflows etc change constantly. Then something as simple as ok I need you to now reorganize these excel columns in this fashion and send it here instead suddenly breaks downstream and upstream steps that were previously solved for. Now I need to spend even more time fixing those items and because I have “successfully” automated something that wasn’t automated before doesn’t mean there is no more to automate.
On the contrary… now the 10 things you couldn’t try to automate before are ready for you to try and guess what? They break frequently too and your back in the token burn cycle. Less intimate projects but way more of them.
Exactly, I work for one of the largest financial institutions in the world and I lead process automation and optimization projects that make this abundantly clear over and over.
After doing this for years now I struggle to see how AI can replace even the most basic roles. One simple context misunderstanding and a loss of millions wipes out most gains that may have been present. Intelligence is one thing but retained context and systems thinking is another.
Example: I know that through various side conversations that even through Bob is this clients rep, I know the client has had some disagreements with him lately. I’m not going to send their follow up information to Bob again I’m going to send it to Greg this time to make sure they feel heard.
Generative LLMs churn out statistically likely sequences of words, pixels, or computer code. That’s what they do, and it’s all they do.
That isn’t useless, by a long shot. Quite a bit can be accomplished by generating statistically likely but unpredictable sequences of words or pixels or computer code. Yet the gap between this and the expectations that have been loaded on the AI phenomenon is at least as significant as the actual potentials of the technology. Listen to the promoters and true believers of AI hype and you can count on hearing claims that LLMs will be conscious, self-aware beings in a matter of months, that they will soon be far more intelligent than human beings, and that once this happens AIs will usher in Utopia if they don’t decide to exterminate us all first.
In point of fact, all of this is hogwash. Outside of a narrow range of uses, generative LLMs perform very poorly. Loud predictions of vast economic payoffs to companies that adopted them have turned out to be so much hot air, and the best LLMs on the market still perform very poorly indeed when confronted with genuinely new challenges—the sort of thing that human intelligence takes easily in its stride. A recent study, one of many, thus showed that LLMs can handle less than 3% of business tasks in the real world.
@ArvindKampli@dayonefoundry Yes! I can ship more of whatever is useful to my coworkers but now instead of maintaining 4/5 apps, dashboards or whatever I need to maintain 9/10. In the end it’s a wash and I have a different more annoying kind of stress as you said. More to keep track of…