If I wanted to understand AI from the ground up, these are the accounts I would start with: @jun_song local open models and AI hardware @0xSero making local AI faster and cheaper @Da7_Tech honest takes after long model runs @haider1 honest takes on AI labs and what makes models work well @MiaAI_lab running models on Spark @morganlinton coding agent benchmarks and which model to use @kimmonismus every AI news with good context @0x_kaize agent loops and model rankings @notjazii tracking model leaks and testing them @Teknium everything you wanna learn about Hermes agent

this guy is seriously one of the most genuine people in the AI space, so if he’s telling you grok bot is the future, I would pay attention.. I think once you actually give grok bot a proper try, you will realize how good it already is for something this early it’s obviously not perfect yet, but you start noticing all these small details that make you see how huge the potential is I’ve already been able to automate a bunch of stuff with it from tracking my payments and scanning news / markets etc to handling simple everyday tasks I would normally have to do myself.. in short, what will stand out the most is the user experience.. you will probably enjoy using it more than anything else.

pretty much everyone i know who never touched openclaw or hermes is enjoying grok bot and i think i get why it’s simply because most people don’t care about agents, tools, environments or any of that stuff they just wanna tell it what they need and have it figure out the rest that’s what grok bot gets right you don’t feel like you’re building an agent but it’s like giving something a job and that’s a way easier idea to sell to everyone

prediction: grok takes the #1# spot in 2027 it’s really getting hard not to bet on grok/xAI grok bot made me realize that…imagine grok 4.7 powering those agents > high level coding > deep research, X search + distribution X gives grok insane real time context, and grok bot makes searching and using all of it feel ridiculously natural.. so no matter how good openai or anthropic get in the coding benchmarks without an ecosystem that keeps pulling you back in, they’re missing a huge part of it..

prediction: grok takes the #1# spot in 2027 it’s really getting hard not to bet on grok/xAI grok bot made me realize that…imagine grok 4.7 powering those agents > high level coding > deep research, X search + distribution X gives grok insane real time context, and grok bot makes searching and using all of it feel ridiculously natural.. so no matter how good openai or anthropic get in the coding benchmarks without an ecosystem that keeps pulling you back in, they’re missing a huge part of it..

it’s clearly a compute problem, as tibo is saying here, but they could make it much better.. If you don’t use much compute for a while, you build up extra usage..then when you need a long coding session, you can use what you saved sure, it’s the same compute limits overall, but way more flexible depending on how you actually use the product, so your compute is there when you actually need it. well I hope we don’t see the 5 hours limit at all again… but this could be a good idea

it’s clearly a compute problem, as tibo is saying here, but they could make it much better.. If you don’t use much compute for a while, you build up extra usage..then when you need a long coding session, you can use what you saved sure, it’s the same compute limits overall, but way more flexible depending on how you actually use the product, so your compute is there when you actually need it. well I hope we don’t see the 5 hours limit at all again… but this could be a good idea

i think what’s interesting about grok bot is that it seems to be starting with the most universal stuff first.. research, images, spreadsheets, planning… basically the stuff almost everyone actually needs.. and it’s just working, people seem really happy with the results if they can keep adding depth without adding complexity, grok bot could get ridiculously powerful fast that might be the smartest part of the UX… start broad enough that anyone gets value immediately, then let the specialization happen naturally over time instead of making users build an agent system, let the agent system grow around the user… to me that feels like a much more natural way to get normal people into agents without ever having to teach them what an agent even is…

i think what’s interesting about grok bot is that it seems to be starting with the most universal stuff first.. research, images, spreadsheets, planning… basically the stuff almost everyone actually needs.. and it’s just working, people seem really happy with the results if they can keep adding depth without adding complexity, grok bot could get ridiculously powerful fast that might be the smartest part of the UX… start broad enough that anyone gets value immediately, then let the specialization happen naturally over time instead of making users build an agent system, let the agent system grow around the user… to me that feels like a much more natural way to get normal people into agents without ever having to teach them what an agent even is…

hermes is the best harness i’ve ever used… like the tool calling alone is ridiculously good pretty much every llm i’ve plugged into it just feels better it made me realize how much of what we think is model performance actually comes down to the harness.

another day, another reminder that unsubscribing from Anthropic was the right call like what kind of response is this? you guys have been silent for weeks while people complained about Opus 5 and now the response is basically that we’re just not using it right? apparently we were supposed to give it CPU metrics and let it hillclimb for hours lol most people just want the model to follow instructions, make the right changes, stay on task, and not overcomplicate simple things. if customers have been telling you for weeks that the experience got worse, then maybe explain how we’re actually supposed to use Opus 5. show people what it’s good at, show the workflows, show how to get the best out of it. staying silent for weeks and then basically saying “people are sleeping on it” is such a weird response when your own customers have been telling you exactly what feels worse..

I’ve been testing Grok Bot for over a week now, and here’s what I think Grok Bot would be way better if > cursor pro usage actually lasted the whole week ..the current limits run out way too fast > grok heavy and cursor ultra shared one pool so unused usage should not go to waste. > you could talk to agents live real back-and-forth not only voice memos that we have now > every bot had its own spend cap > agents had separate logins and files so one agent should not affect another > there were audit logs, so you can see what every bot did. > you could pick the model you want for the job.. honestly, just adding these would make Grok Bot 10× better.. well, it’s not bad at all for something that came out like two weeks ago.. It’s just that the reliability, usage, and agent controls still need some work..

I’ve been testing Grok Bot for over a week now, and here’s what I think Grok Bot would be way better if > cursor pro usage actually lasted the whole week ..the current limits run out way too fast > grok heavy and cursor ultra shared one pool so unused usage should not go to waste. > you could talk to agents live real back-and-forth not only voice memos that we have now > every bot had its own spend cap > agents had separate logins and files so one agent should not affect another > there were audit logs, so you can see what every bot did. > you could pick the model you want for the job.. honestly, just adding these would make Grok Bot 10× better.. well, it’s not bad at all for something that came out like two weeks ago.. It’s just that the reliability, usage, and agent controls still need some work..

so I’ve used Grok Bot enough, and it made me realize what Hermes is still missing... > Grok Bot feels super smooth for non-technical users... you just open it, start chatting, and spawn more agents whenever you need them. almost zero setup, and nothing is really that complicated... > Hermes is still amazing UX-wise, but it can feel a bit too advanced for beginners... it’s easy if you’re technical, but not really for everyone else. well i think Hermes is already moving in the right direction... they have bot mode and Nous Portal handles models and tools under one login, and computer use is already there... but it still doesn’t package everything especially an always-on cloud computer into one simple “pay, open, and start” experience.. I think Hermes desktop should have two modes: simple: one subscription, managed cloud computer, computer use, and setup.. all handled in one place... so you just pay, open the app, and start using it. advanced: full control to choose and customize your own models, VPS, keys, tools, setup, and everything else... that’s where I think Hermes should focus... Hermes already has the advanced features, it just needs a truly simple mode. some people would genuinely pay just because everything is handled in one place, and you’ve gotta give them that option. @NousResearch @Teknium

so I’ve used Grok Bot enough, and it made me realize what Hermes is still missing... > Grok Bot feels super smooth for non-technical users... you just open it, start chatting, and spawn more agents whenever you need them. almost zero setup, and nothing is really that complicated... > Hermes is still amazing UX-wise, but it can feel a bit too advanced for beginners... it’s easy if you’re technical, but not really for everyone else. well i think Hermes is already moving in the right direction... they have bot mode and Nous Portal handles models and tools under one login, and computer use is already there... but it still doesn’t package everything especially an always-on cloud computer into one simple “pay, open, and start” experience.. I think Hermes desktop should have two modes: simple: one subscription, managed cloud computer, computer use, and setup.. all handled in one place... so you just pay, open the app, and start using it. advanced: full control to choose and customize your own models, VPS, keys, tools, setup, and everything else... that’s where I think Hermes should focus... Hermes already has the advanced features, it just needs a truly simple mode. some people would genuinely pay just because everything is handled in one place, and you’ve gotta give them that option. @NousResearch @Teknium

wild, now i would say its pretty worth it. so, you no longer need to pay $200–$300 to try Grok Bot. It’s now available with: SuperGrok Plus: $100/month Cursor Pro+: $60/month Cursor Teams: $40/user/month i think give it a try, theres a limited free trial then decide if you wanna pay $60 or not...

why do I have a feeling Ox Alpha is an unreleased multimodal GLM model >1M context window >tokenizer behavior >and API profile all look way too close to GLM-5.3. it also even introduced itself as a Zai model while testing it since the public GLM-5.3 is text-only, my guess is this is the upcoming multimodal version being stress-tested before release.. i mean, i might be completely off, but everything lines up a little too perfectly for this to be random..