namredips

[namredips @19_namredips]

@TomBarman94 @aiwithadb Not an unfair comparison at all really… probably a difference in need or demand. But still a fair potential outcome. I wouldn’t put it all on technical advancement, but I would certainly assume a variable

[namredips @19_namredips]

@ABasapathy @aiwithadb Everyone will be lean like this… ai doesn’t drive its self. You will need more people to compete because using AI will be just as ubiquitous as using a computer. Success will continue to be in the agency if others, even if there is a dip for a while.

[namredips @19_namredips]

Lots!! Continued work on a major saas app, work on my rust “all agentic company” interface. Finalizing a better code reviews that’s more deterministic than not and if time permits the AI lab I have been building in in rust to experiment with fine tuning LLM’s, training CTRNN’s and experimenting with both working together in an agent. The last being the fun one

[namredips @19_namredips]

@satsuki2694 So this is tok/s, where do I find quality of work… say coding for each of these… I assume they are not full 16. Even relative benchmarks would be great…

[namredips @19_namredips]

@trikcode Time to start building agent friendly services, ride the wave as people jump. Probably won’t even know the jump. “Hey muse can you find me a good deal on some new dishwashing liquid… “ muse ignores Amazon and finds it at an agent friendly store.

[namredips @19_namredips]

@AaronOrendorff Let’s say you are building a web app of some sort. It’s got multiple types of users. You can handoff a set of user tests by persona to the lead. It can then divide up by persona the tests… use a browser like a real user and collect results. Claude disrupts your laptop

[namredips @19_namredips]

@thsottiaux @rl_env Also quick to judge period, it’s in no way clear that “normies” want just one agent.

[namredips @19_namredips]

@spacejohnston76 @Aditya_181105 Oh I did forget about this one!! Maybe on purpose!!

[namredips @19_namredips]

@TechnPrep @Aditya_181105 I came looke for this!!!

[namredips @19_namredips]

@DanielMiessler It’s an efficiency thing

[namredips @19_namredips]

It’s a mistake to think that when using AI the writer is the reviewer, unless you are using the same session that wrote the code to do the review. Further, I would bet good money that the reviews the AI does is better than 90% of reviews for the past 30 years… reviews have always been subjective. Tests are deterministic and the real answer to “is the code good”.

[namredips @19_namredips]

@ananyairl Shhhh….

[namredips @19_namredips]

@nmPraveen @thsottiaux Dude calm down… it’s not that big of a deal.

[namredips @19_namredips]

Wait till November when the current $200 is half usage than it is today. Not saying people will pay it, but it’s likely the $500 will seem like the old $200. Hard to say… wonder if the current $500 will actually be less. Since it’s 25x of something that will be less than before… might have that misunderstood

[namredips @19_namredips]

@lxe I seriously can’t tell if this is sarcasm or not… given no one needed to reverse engineer it. Maybe if I click in the link it would explain… but posts like this generally are a bad idea to click on the link

[namredips @19_namredips]

@ElitzaVasileva 185 issues, optimized workflow. Plan, review plan, implement tests, implement/test loop, code review/fix loop, deploy/test loop, UAT/test loop. Plan => deploy/test per issue. This doesn’t even include the creation of the issue. Which is a much slower back and forth with the model

[namredips @19_namredips]

@Teknium Man I love Hermes, but it needs a better TUI. I just can’t make the current work for me. I guess I will have to make my own. The desktop app is ok… but desktop coding is not great. I need something the orchestrate across multiple configure at least…

[namredips @19_namredips]

@JeudryPena28376 Uhh… ultra got the reset as well

[namredips @19_namredips]

@thsottiaux Also… I hope this applies to the cli!!

[namredips @19_namredips]

@thsottiaux That’s good news. Already one of the more efficient models and the best “coder” for the languages I am using. After next reset, i guess i will be able to see if I notice any difference. 👍