4 ms·

If someone from Grok is reading, don't waste time on these chaff features. The market will eventually deliver better 3rd party solutions to all of these things.

by Cakez0r 5mo ago

If someone from Grok is reading, don't waste time on these chaff features. The market will eventually deliver better 3rd party solutions to all of these things. There is an audience that isn't interested in these walled garden features and are only interested on intelligence per dollar.

torginus 5mo ago
Aren't they 'wasting' time on these features exactly because the engineering requires a different, more traditional skillset from the ML work model people do, and can be done in parallel?
raincole 5mo ago
Lol I wonder when Anthropic discussed the idea of Claude Code internally, were there bozos saying "3rd parties will eventually deliver this so we shouldn't waste time one it."
- Cakez0r 5mo ago
  Power users are hotswapping these models into their own agents (hermes, openclaw, etc) which have their own systems for project management, memory, interacting with tools, etc. The important metric is intelligence per dollar. Can I drop this model into my harness and have it be cheaper without losing intelligence. That is where the puck is heading.
- wincy 5mo ago
  Personally, my work doesn’t want to get locked into a single LLM provider so we use Cursor. Much easier to fight the big corp software approval battle once then switch around the LLMs to the new hotness (provided legal has the requisite data sharing agreements in place, we’re not supposed to use Chinese models or Grok) but I can switch between Anthropic and OpenAI models at will.
- wyre 5mo ago
  The only good thing Claude Code did was bring coding harnesses to a wider audience. It is not a good harness.
  - jmalicki 5mo ago
    What are good harnesses? I haven't yet been able to get good agent teaming approaches out of other harnesses yet, before that feature I mostly regarded the space as competitive, but until another harness can do as well with Claude models it seems like it's better for now?

by Cakez0r 5mo ago

torginus 5mo ago

Aren't they 'wasting' time on these features exactly because the engineering requires a different, more traditional skillset from the ML work model people do, and can be done in parallel?

raincole 5mo ago

Lol I wonder when Anthropic discussed the idea of Claude Code internally, were there bozos saying "3rd parties will eventually deliver this so we shouldn't waste time one it."

Cakez0r 5mo ago
Power users are hotswapping these models into their own agents (hermes, openclaw, etc) which have their own systems for project management, memory, interacting with tools, etc. The important metric is intelligence per dollar. Can I drop this model into my harness and have it be cheaper without losing intelligence. That is where the puck is heading.
wincy 5mo ago
Personally, my work doesn’t want to get locked into a single LLM provider so we use Cursor. Much easier to fight the big corp software approval battle once then switch around the LLMs to the new hotness (provided legal has the requisite data sharing agreements in place, we’re not supposed to use Chinese models or Grok) but I can switch between Anthropic and OpenAI models at will.
wyre 5mo ago
The only good thing Claude Code did was bring coding harnesses to a wider audience. It is not a good harness.
- jmalicki 5mo ago
  What are good harnesses? I haven't yet been able to get good agent teaming approaches out of other harnesses yet, before that feature I mostly regarded the space as competitive, but until another harness can do as well with Claude models it seems like it's better for now?