ActionHank

I am running it locally at home with tailscale for access outside of my home network. I added a tavily key for search and pageloads. There are various FE projects that can connect to Hermes, they also have a frontend, but it's not a "search" interface.

If you're looking for something more typical of search + AI, DDG is probably the next best. If you're looking for a search assistant that will find you a direct link to the leaked gta 6 footage that people are talking about or to summarise a comparison of 2 oss libs, Hermes is probably the one you're after.

DDG allows you to turn off ads.

Everyone I mentioned it to scoffed at the idea of paying for search.

I tried Kagi for a couple of months and then cancelled.

I really liked the AI search, but it burns through credits too quickly. I have since switched to using Hermes, it's not quite as good yet, but it's a fraction of the cost.

I liked the idea of ad free, slop free search. There were some problems with that though, which were hard to push past. I really like being able to search for stores and local restaurants. Maps integrations were hit or miss, mostly miss. It was also weirdly slow for me, even to load the search page without a query.

Final point was their published and then unpublished support of Yandex. The CEO tried to argue some "search without borders" BS, but really they are taking your money that and sending some of it to a Russian company which is either directly or indirectly funding a pointless war. I took issue with that and remembered it every single time I opened the page. In particular it really didn't sit well with me that the response was flippant and then they took down their list of sources.

I've switched to DDG and have been happier, YMMV.

“Author” is real stretch, “Prompter” is a better fit I think.

Hermes + GLM 5.3 Flash for orchestration. omp + Deepseek 4.1 Flash for coding. Occasional call out to claude for a review on a large change.

Hermes Kanban works stupidly well here and omp running on a slice in a worktree fans out really well.

Wild that this is the first question I ask at this point. After that it’s a question of how it stacks intelligence wise vs glm or ds.

OAI / Ant are in for a reckoning.

No, I just know how business works. When you are doing something like this you establish the bare minimum as the baseline, get your foot in the door, then start testing boundaries.

Also these buildings will be abandoned within the next couple years as models move more towards the edge.

I'm saying that just because they say that they will generate their own electricity, or require less water than a farm, doesn't mean that they will continue to do so into the future.

If you believe all of those counter points will hold true for even a month after operation starts you're real naive. People driving the plans to build out this infrastructure do not care about a single point listed there and will lie to get it over the line.

You know who has a lot of power / space / water?

Private compounds owned by billionaires who believe data centers are the best. Seems to me, they already have a solution to their problem.

I for one think it's pretty cool how far too many people bend the truth or outright lie in tech these days.

There is going to be a reckoning for tech in the near future, because people will remember the lies and it's gonna be hard to beat the history.

"Have a text area"

You've already failed. Config files are for more than just editing in an environment that you solely control.

And also, because no one seems to see it, negative profit for the foreseeable future whilst competitors have locked in "good enough" and a fraction of the price.

Did you even read it? Wild the fanboyism for AI has rotted out the brains of so many people. It's a tool, it's useful, people will adapt.

The reaction was from the strip mining of pending human discovery, mathematicians were already using LLMs from frontier labs who then published suspiciously similar solutions or found connections that were just waiting to be discovered.

Gotta be honest, any time there's a solution like there that makes it more "human" they've just removed the structure / syntax.

I'm guessing so you don't need to learn them, so then anyone off the street can edit the files, but it will break so bad without those guardrails, usually in some non-obvious way.

Leading us back to inventing some new "human" thing, but with more implicit rules, and the cycle continues.

No, not when that business spends more than it makes to bring in that revenue.

A better way to phrase your thinking might be that revenue is an indicator of demand.

The issue here is that they've not really tested what happens to demand when the prices moves to where it needs to be to cover the cost of keeping the lights on.

"Just don't look at those cheap af Chinese models pls"

There is no company on earth that can run on customers being interested alone, you also need to offer a product they can afford at a prices that generates profit.

Right now, I could start a business selling gold bars for $1 a piece. There would be more people interested in that than Anthropic.

Can we all please stop acting like revenue is somehow an indicator of a successful business?

Are you one of those people who lost half their money to spacex? You sound like you are.

Because it is, there isn’t enough money to be made to cover their valuation and their competition are essentially free by comparison.

I’ve discussed with many people at work, all engineers, they’re weirdly oblivious. Most are shocked when I say things like OpenAI and Anthropic are gonna crater.

Categorically, if you ship more bugs per engineer, you fix more bugs per engineer.

I suspect that once the hype finally settles we will see a drop in demand. Turns out Tina from HR didn't need that vibed out custom calculator webpage after all.

No, take a look into people who bought used GPUs that were from crypto farms. It's not just thermal paste, the chips become unreliable because of degradation of the circuits etched on the chips themselves.

Thats the weird thing, scarcity drives value up so long as there is scarcity.

Once there is enough, or even a glut, it's over.

Yeah, and anyone who had their crypto there will move elsewhere, then Binance has -12b

Crazy thought, but what if they went via the Claude Code trojan?

But with Dario's goofy expressions, gonna be *chef's kiss*