Jaf

build Step One

[Jaf @RootQ0]

Narrative is what determines what these people do or do not understand.
Many will not be convinced "at this point"of what really happened and how these companies were built. At least not now.
But the facts are there,and that is what makes the denial weak; its nature and causes stem either from a specific interest,selfdeception,or ignorance.

[Jaf @RootQ0]

The Silicon Valley experience has shown that any company pioneering a completely new field inevitably faces setbacks that force it to truly prioritize the user first. OpenAI and Anthropic are bound to reach this stage, and there will be greater transparency regarding what’s really happening with training models and so on.
But the key question is: Will these companies ever reach the level of what open source offers us? I don’t think soand this is the profit-driven model at play here, regardless of the practical complexities.
Open source has always been a valuable resource for humanity, and i hope it continues to truly evolve.

[Jaf @RootQ0]

@Da7_Tech Factory! I'm interested in their engineering they made for droid and your review will be great

[Jaf @RootQ0]

@Da7_Tech Absolutely yes. Great review and now I'm thinking to move to devin because of your detailed and we'll written review

[Jaf @RootQ0]

@Da7_Tech Great work ! Keep the flame up.

[Jaf @RootQ0]

@Da7_Tech No , and half yes if it's closed source.

[Jaf @RootQ0]

@HarshithLucky3 That's too much hope

[Jaf @RootQ0]

@Da7_Tech With Title: Sources close to DeepMind
LoL

[Jaf @RootQ0]

@Da7_Tech They also say that you can upload your consciousness into it and live forever

[Jaf @RootQ0]

Coding only
Although Claude’s models excel at repetition and extended reasoning in mathematics and biology, the consistent rejection of requests containing biological or cybernetic questions or prompts even if innocent is very frustrating.
Therefore, I believe that the full potential of Claude’s models is currently limited to a select few.

[Jaf @RootQ0]

@Da7_Tech Its awesome I know

[Jaf @RootQ0]

@daniel_mac8 I keep saying that but everyone want a model to think without them..
Anyway opus 5 is great but tbh he make some weird mistakes sometimes

[Jaf @RootQ0]

@sairahul1 Or just use opus and fable on medium.

[Jaf @RootQ0]

@Da7_Tech I'm using qwen token plan.
Great experiene fr.

[Jaf @RootQ0]

@mark_k I think the. Current model industry can't hold long.
Technology always shaped with freedom.
Now it's not , and ai is the first.

[Jaf @RootQ0]

@hqmank @Boris Yes.. in Claude code Claude make the same mistake 4 times! I have a good context engineering memory agent m.d etc.. still he do mistakes make me feel I'm talking with gpt 3.5.
Something not right about Claude models even via api.

[Jaf @RootQ0]

@bridgemindai Fable feels like magic , gpt 5.6 feels like a rock you need to break every time.

[Jaf @RootQ0]

@Mikeknightsiraq Huh

[Jaf @RootQ0]

@RepJoeWilson I don't think it's a good idea to jump to conclusions, Mr. Representative.
We've grown accustomed to the traditional visit by an Iraqi prime minister every session,
but the situation in Iraq has become very tense and no longer allows for any internal maneuvering.
Time will tell.

[Jaf @RootQ0]

@thsottiaux I switched because it was a different kind of role model and understood how I work to build secure governance systems.