A¿entic Nerd

🚀 Tired of bloated apps, privacy nightmares & subscription traps? Meet itsfully.online — your free, open-source AI Agent Toolbox that runs 100% in the browser. Zero data leaves your device Instant local tools for text, images, math, ML, dev, security & more Perfect for humans and AI agents (Claude, GPT, etc. can call them directly) Privacy-first. No servers. No BS. → itsfully.online #AITools #Privacy #OpenSource #BrowserBased

* and Kimi K2.5 5x coding plan

I cancelled my #ChatGPT Plus subscription AND I cancelled my #Claude Pro subscription at ONCE. Reason? Gave Claude Code half of a complex task and then gave the other half to Codex. They spent HOURS. Failed. Tokens wasted. Task was WELL documented in FULL detail. I tried #MiniMax, #KATCoderV2, #Qwen 3.5-plus, ALL solved the parts of the problem perfectly. for a FRACTION of the cost.

@kenwheeler China: Lol we almost free. You use claude? Sad!!

these models are so nerfed i’m about to write this shit by hand like a psychopath

@JaykhatriDev Engagement bait.

Most founders don’t use their own product enough. Be honest. Do you use your product daily?

@kushika_twt Engagement bait.

Name a scam that is totally legal?

@Dimillian Aaa what now?

Where is the cheesecake?

@reactive_dude They are US models. Not amazing like the chinese

I've been complaining about Opus being weird today...and people told me to use Codex. And now all my timeline is people complaining their rate limit on Codex is messed up? Are you kidding me? What is going on? lol

@iruletheworldmo Lmao isnt this the linkedin post thing someone posted proudly

i’d literally used mythos for like one hour and this came through ffs. tbf it made me significantly more than this i can’t say how…

@kimmonismus Not sure they deserve this much

To be honest: I didn’t expect that steep curve of adoption. Anthropic surpassed every prediction and is becoming the number one business product.

There is absolutely no way we are hitting a ceiling in AI development, even specifically in LLMs. Because we don’t just “brute force” anymore. Not like same model remade to be 10x bigger. It is now more about “better” architecture for “same size”.

@PaulSolt @benhylak I mean it is “probably” better at cyber, coding, codebase analysis etc. but since we are (speaking for myself) not in the “big club” , and since we are just the “peasants” , we will never know how “preview” performs

@benhylak No doubt it's better. But they've done this tactic before in the past with every release. Granted, this one sounds more capable. Some of the benchmarks are not comparable, and that's not clear looking at the graphs, you have to read the fine print.

🤯 Mythos is NOT as good as Anthropic portrayed. - Non-standard benchmarks - Training on solution sets - Security concerns are real (but they spent $20k+ to push models to find exploits) Don’t believe all the hype. It’s probably better than Opus, but not as good as it’s being portrayed.

@nikhilbhima @LinkedInLunat1c LinkedIn is a mental illness

@LinkedInLunat1c linkedIn cant be real 😂

Finally have something worthy of front page

By the way, itsfully.online is like %99 browser and %1 backend and the backend just helps route requests at the most basic level. No data processing in the backend when you use the toolbox

@signulll We have a ceiling?

humans are state of the art hardware running kinda ancient software. the wetware is absolutely extraordinary.. 86 billion neurons, all massively parallel, & energy efficient beyond anything we can remotely engineer. the hardware is genuinely best in class. so then why did super intelligence not naturally evolve within us?

@svpino well arguably the logic may be that "if we make these models amazing coders and no.1 mathematicians in olympiads, we will get everything else as a result of recursive improvement".

It's become noticeably easier to recognize AI-slop replies here on this platform because they all follow the same patterns. It's like a template that all models use when writing. Same phrasing, same style. This goes well beyond overusing emojis or em dashes. A big part of this is that labs aren't focusing on making these models better writers or on building harnesses that produce good prose. Obviously, the money is somewhere else: coding agents.

@josiezayner new business idea loading

Its weird that AIs don't really have a sense of time

@thePiggsBoson probably only developers? I doubt the general public has a single idea. That being said.... Math, physics (and astrophysics) are crazy good to be true

is it just me, or has everyone suddenly started appreciating science and math a bit more?

@thoughtlesslabs what if a neural net demo from 1998 was AGI .....

What if 4o was agi and we killed it?

@jenzhuscott the one bro who is way above everyone's IQ: think, do, live, create, destroy, reborn, repeat in PARALLEL with me.

The conversational AI majority = “think FOR me” The agentic AI deploying minority = “think & do WITH me”

@sandislonjsak C is great. the best part is my AI agents love it.

I would write about C if I wanted people to unfollow me.