Claude code is good, Opus as general model is a hot hallucinating piece of garbage. I asked it to review a single page 50 rows, 2 column excel file. It hallucinated almost everything in the file. It repeated that for the next two files I asked it to review, these were tiny files, barely 20 kb.
The funny part is Opus was the one which generated the files in the first place.
This was Opus 4.7 High. So no thank you, Anthropic.
TPU8t is for training. But even still, once you’ve trained, you need to run the model too. And these kinds of models already have a huge latency hit so there’s not much hurting running it away from the trading switches.
I considered going to hetzner at one point but I read a lot of stories around hetzner that didn't inspire confidence. Primarily that they're not really that much cheaper than going to other companies offering something similar.
If some people can chime in with their positive experiences I might switch.
I shudder when I read DRM after building a embedded display raspberry pi build. There was a problem with Raspberry pi kernel incorrectly detected an HDMI display was connected even though I was trying to display to DSI. The only work around I had was to write a kernel module that would disable HDMI as a possible interface.
The problem is how bullshit transactional emails are when you're outside of AWS. If you're not expecting to use 10,000 emails a month but would like the option to go over the free tier without committing to 10,000 more. Just let me pay per use FFS.
The reason AWS does that is because there is a lot of base level work to verify you as "not a spammer" and to keep verifying you. So this is their way of making sure you pay the base cost.
They could price per use, but it would have to start with a base fee that is about the same at 10,000 emails.
The base fee doesn't need to be monthly, they can take $50 dollars as a one time registration verification fee. That should be enough to cover their compute costs for the year, specially when you pool in that from multiple customers. Who is spamming with 100 monthly emails? How much compute do they need to verify you aren't spamming. They can bake all of that into the pay per use price, they choose not to and I'm glad cloudflare is offering this.
Maintaining the reputation of an IP address is the issue here. If one bad actor sends just a few emails that get marked as spam, the entire IP gets marked as bad. That's basically what you're paying for.
Also, the person who just wants to send a few 100 emails a month is actually far more likely to be a spammer. So it's also a way for them to eliminate those folks.
And lastly, the support burden can be high.
AWS has basically said they only want serious customers, let the other guys worry about the small senders.
I agree, software (software startups) has always been the golden child of investors because of how cheap it is compared to hardware or any other physical good.
Good software is expensive regardless of the involvement of LLMs because you need someone to take responsibility. Large companies will save a buck because there may be fewer people needed to take said responsibility, but it's probably a marginal saving compared to the overall scheme of things.
I literally said this exact thing verbatim with someone who was worshipping ai for code. This is an insane idea, really cool.
One bit of feedback, I'd love if this could be a little more aesthetic in the visualization. The method names are a little hard to read, white on light green.
It was just the first look...My primary focus on was on the Engine. I will upgrade the UI and will surely change the UI you mentioned. You can also contribute if you wish as it is open source (built in nextjs)
The funny part is Opus was the one which generated the files in the first place. This was Opus 4.7 High. So no thank you, Anthropic.