Some questions...
1. How do we unlock the massive upside of AI while avoiding a dopamine drip induced algorithmic oppression that sucks all joy out of life?
Just tested positive for Lyme after weeks of unpleasant symptoms.
Hit me with your best unconventional protocols for fast recovery. (Will stack on conventional.)
@jimjannuzzio Is there a best by or expiration date? 241 days is a long time.
How can someone be so disorganized to leave $150k in inventory abandoned???
@Muse Question re the referral bonus. I invited my wife last week and she started account creation but wasn't allowed to join. She was just given access today but it says that she is beyond the 48 hour window to add the code?
The thing that has me reaching for Muse for an increasing amount of my personal (i.e. non business and work) agent use is the layout.
I think they are closer to solving thread organization than anybody else. I still think it could be a lot better, but what they have is already preferable to the clutter I encounter with every other tool.
I especially think the feed and ideas sections are valuable, although I wish conversations nested better.
@DanaPeters0321 That's awesome. How much did you invest upfront to be able to build that? Did you do a test run and let the market pull this out of you or did you yolo a huge batch?
@ryantalks1024 One funny thing is that it appears to have ruined them simply by using them correctly. But since 95% of people do not use them correctly, they now seem overused and out of place.
One potentially positive outcome of this is more dispersal.
A negative unintended consequence of the internet was specific places trending as the "hot" spot. Not necessarily because it was truly better, but because it had captured the imagination of the travel blogs (and later tiktok).
An example: I was recently in NYC and running my old playbook for generating Alpha on the best bagel shop. But then I stumbled on a local New Yorker lamenting how the internet has been missing the point on great NY bagels. The magical thing is that you are never more than 5 minutes from a bagel better than you can get anywhere in the world. Trying to find the best one may actually be destroying the Alpha, as you end up spending a lot of time and energy in the hunt when you could have just stumbled on some fun, random spot, enjoyed the bagel, and moved on with your day.
How this relates to your original post...
I think it is possible agents end up doing a better job at helping people find alpha that was hidden from them because they are much better at sorting through large volumes of details to find places that you uniquely will appreciate, either because of random personality quirks or because it knows where you waste time and can help you not do that or because of some other reason entirely.
@CLEMENTAIEXPERT I'm confused - this seems like an architectural recommendation directed at someone building a retrieval system... or am I reading this wrong?
Agents don't search the same way humans do.
For one, they can review a lot more information. Which means the coveted top spots on Google may not matter as much as they used to.
This can cut both ways though. One of the businesses I'm helping right now is an HVAC shop, and I recently did an audit of how they show up in organic AI / agent queries related to their services.
I found out that they aren't being displayed by any provider. For each query and each agent, I kept asking why it made the recommendations it did and to give me a list of additional businesses it considered, along with the reasons it ultimately chose not to include it in results.
For the business I'm working with, a consistent finding was that it's lack of reviews on Yelp were sinking it. In this particular market, Yelp is not a place people are using to find HVAC help. Main competitors have <15 reviews total as well. Yelp doesn't show up on the first page of search results, but the agents are seeing that this business has only 4 reviews with one calling the owner an a$$hole, so they decided not to recommend it at all.
The main takeaway right now is just that it seems likely that agents are going to process a lot more information before making a decision than the average person, and there are implications to this. Worth thinking about more.
Fair point. Low consumer confidence combined with high inflation since last replacement generates sticker shock and delayed replacements, leading to lower install revenues and higher spare capacity among install technicians leading to a need to cut margins to try to secure more jobs.
@farrowhardy1 What makes right now worse than any other time? This seems like something that is always part of the market, i.e. the bottom performers gasping for air and doing anything to generate cashflow...
or are you saying you are seeing more of it now?
Can confirm, running a farm is expensive...
But $500k a year is crazy just to raise family food. Is there no revenue generated off of this? Where is all that money going?
You could pay a farm manager a 6 figure salary and run an entire 100 head cow operation with winter feeding and still have hundreds of thousands left over!
Thank you for sharing. As an American, I share many of the same general feelings and questions.
Like you, I find myself rooting for them to shift the Overton window and curious about what alternatives become possible in the wake of that.
I also find myself wondering what exactly it means for everyone else if their thesis is right. Is there a version of this that maintains (most of ) the value while being something a regular family can actually do? I understand that experiments have been run with low-income families, but that is very different from the model being accessible for the average child.
My money is on Muse. Not necessarily for the X crowd, but at this point that is where my money would be for the biggest general market share in 12 months. Fast, good enough, much simpler, already integrated into existing apps people use... and pretty much free.
Plus aggressive marketing. And an insane budget for talent.
Introducing Claude Opus 5.5, the first model in our new Claude 5.5 family.
It performs at the level of Claude Fable 5.1 for most tasks, and costs 40% less to run than Opus 5.