TLDR: GrokBot great for simple tasks and users, bad for power users. Hermes good for power users, awful for average users. Openclaw is unusable garbage. The transfer to Grok @bot ultimately didn't work out for the full agent stack. It's too complex for it. Was running into consistent problems with it getting confused on which tools to use, who handles what, etc. I attempted a @NousResearch transfer, but that went horribly. Burned up 3 full GPT resets in a day, finally got it going, and the employees hated it. Everything was full of word vomit, tool calls, error notifications, etc. It's simply not made for the average person, even on the user side. So I rebuilt the whole thing from the ground up, with knowledge preserved in Obsidian. There's now only one agent that employees work with on a day to day basis with only the tools and knowledge needed for them. That stack still runs on GrokBot, but I had to wipe it and start fresh entirely. It actually runs very well for what they need it for, but the $200 Ultra Plan is hitting limits and charging overages. Wasn't an issue at all running the entire stack with Codex, which was much more complex and running constantly. The rest of the agents will work off the same Obsidian vault for company knowledge, but no longer be used for day to day operations. Instead, I'm rebuilding the management aspect of the business separately. Probably with Hermes and Codex, since the word vomit doesn't bother me as much.

ai agent/harness tiers

Free access to Instinct & Muse feels like $3 Uber Pool rides all over again. Enjoy this stage of the subsidy cycle, we'll be nostalgic for it later 🫡

day 1 observations for grok 4.7 ignore the reports that say “it’s terrible” and the only thing they reference is a public benchmark. the same benchmarks told us opus 5 was better that fable - they are useless also ignore the reports that compare models with 3d games - that’s not real work. it's made for attention on social media i used grok 4.7 for a whole day as my firstmate, and it has been a really solid model with visible improvements over 4.5 (i'm ignoring 4.6 because 4.5 has been working better in my experience) key differences with 4.7 - 1. it follows system prompt very, very closely i noticed firstmate showing many new behaviors that i've never seen before, such as asking me to name specific red CI checks that i'm ok with bypassing, and refuse a simple "yolo" instruction i traced it and it's indeed how i instructed it in firstmate's system prompt, but none of the other models followed it closely enough to make this behavior visible - grok 4.7 is the first to pick that up there were a few other similar examples as well. so to me this is a clear behavioral difference 2. it's very "stable" if you've used astra then you know what a "spiky" model is. it can have some genius moments but you occasionally also wonder "how could it be so dumb and doesn't get me". grok 4.7 is the opposite of that throughout the whole day so far, i'll be honest i haven't get a "wow this is absolutely genius" moment yet, but grok 4.7 has been very steady with no big surprises. its behavior feels predictable, which does help it gain trust from me quickly 3. it's a conservative model it doesn't like to take actions without asking, and would explicitly say so this is a bit of a double edged sword, because it means i sometimes have to state the obvious "yes i do want that", but in hindsight a lot of those cases are indeed a bit ambiguous and i may not have preferred the model to just move forward without my confirmation 4. it's a bit slower and costs more than 4.5, visibly turns are taking a bit longer and my quota is draining at a visibly faster pace. i haven't quantified exactly where this is coming from yet so overall, i think it's showing some clearly different traits, and i mostly like the changes. i'm going to keep it as my primary firstmate and observe more if you've been using it, what qualitative insights have you gathered from real usage so far?

one of my favorite use cases of these ai agents is to "find" money. basically just do the things and follow the processes these companies set up but make too annoying for normal people to follow.

Flight delayed 7 hours - asked Muse to file for compensation. 5 mins later i had $250 credit in my delta account. it even found and rebooked me a new flight. it just figured everything out. even responded to the support email itself. shit feels like magic.

Grok 4.7 is out It brings improvements over 4.6 and is especially good for: >Software engineering >Long-running knowledge work >Electrical engineering >Legal tasks Pricing is the same as Grok 4.6

One missile, or seven. Same price. a16z's @DefenseInDepth_ on the arsenal the US could have for the price of the one it's buying: a16z.news/p/magazine-depth-a…

College football’s most carefully guarded secret is the amount of money flowing to the players on each roster. So we reached out to 70 sources across the sport. Estimated budgets for all 68 teams: nyti.ms/4xwoPqe

Holy crap POTUS just phoned in @JensenHuang live on stage at All In Summit We will not lose the ai race! And whatever Dario said this weekend won’t stop our progress This made my morning! $NVDA

This would suck, but we will prioritize great service for customers until we can get back on top of things.