Vibecoder

@dishant0406 @aryanlabde Yeah for sure, it started with copy pasting files etc. I'm building a political analysis service for European politics. Grabs text and docs across 300+ sources, processes it all thru multiple LLM layers etc You can check it out here: bb.mccurrach.xyz

@Vibecoding42069 @aryanlabde arduino to claude code is a solid path. hardware teaches you constraints. software gives you scale. the debugging mindset transfers. what are you building now

Most vibe coders are building their 4th product. Still waiting on revenue from the first three. Shipping fast isn’t the hard part anymore. Selling is.

@awilkinson People do say that business favors pathological traits so not surprising

Most insanely successful people I've met are not even remotely introspective. To me this always seemed more indicative of a personality disorder than a strategy for living a happy life. Extremes create great work outcomes, but not longterm personal or familial happiness.

@dishant0406 @aryanlabde Before that I'd built exactly 0 apps, even though I'd wanted to for a long time... Started with Arduino/esp32 actually and then moved onto more sophisticated apps with cursor, and then finally pure Claude code.

@dishant0406 @aryanlabde Exactly, getting 90% done with several different projects enabled me to learn how to get the best out of LLMs, and apply the experience once I knew what I wanted to build...

@observie Any recommendations? I'd also like to get a pick and place machine one day...

If you're serious about soldering, you need a microscope

@chang_defi Lots of LARPers, scammers and grifters in crypto unfortunately

Her salary comes from $ extracted from dumping star atlas, aurory, genopets, stepn, nyan heroes to name a few solana gaming scams Shut the fuck up

@ThomasVanRiet2 Will there be any physics left by then? I presume this is related to the metric origin?

Conjecture: by next fall most theoretical physicists will mainly use paid AI services to solve their research problems.

@Eli_Albrecht Just give up already, only a matter of time, and tbh, I'm sure you've got more to offer the world than poring over legal documents while charging a fortune to do so. Legal services are an enabler, but not productive in and of themselves.

First, these are all basic form documents that a paralegal can do. Second, I have never seen a draft of a legal document that does not have fatal errors (like really detrimental issues) (“it will get better!” I know) Third, it will take less time for me to review and fix if a paralegal or junior associate does this.

@garrytan Nice work Garry, I will have to check it out. I will reiterate my criticism of Garry's list - it's obviously pro tech overlord propaganda imo - you should drop the editorializing and instead just deliver Intel and analysis without bias

GStack is just pulling on latent space inside the models to be sure And this is the worst the models will ever be

@r0ck3t23 Lol, he's just justifying his fantasy to contain China. You've already lost, and if you try and force it, you'll just create a nuclear wasteland...

Palmer Luckey just dismantled the moral argument against autonomous weapons in three sentences. The choice is not between weapons and no weapons. It is between dumb weapons and smart weapons. And if technology can reduce civilian deaths, refusing to use it is the moral failure. Luckey: “I don’t see the moral high ground in making, for example, an anti-vehicle landmine that can’t tell the difference between a Russian tank and a school bus full of kids.” When you refuse to deploy a system that can distinguish between a combatant and a civilian, you are not protecting humanity. You are choosing collateral damage. Keeping algorithms out of the kill chain doesn’t save lives. It costs them. But Luckey didn’t stop at the capability argument. He went straight to the accountability question. Luckey: “If I’m deploying those landmines in a place where they’re able to make a mistake, and they make a mistake, a person needs to be held accountable. And you say, did you use them appropriately? Did you trade the risks appropriately? Does this actually make sense?” The architecture of the kill chain changes. The liability does not. Autonomous systems don’t absorb responsibility. They accelerate execution. The human who authorized the deployment still owns the outcome. Every outcome. The operators who dominate the modern battlefield will scale autonomous capability aggressively while maintaining absolute human ownership over the decision to deploy it. That is the line. And it never moves. And here’s the part that makes this existential. If you demand a human in the loop for every single physical execution, you are engineering your own vulnerability. Luckey: “The answer is not to say that you cannot have an autonomous air defense system, because if you do that, all someone has to do is they break communications or break your command and control… and now your enemy can pummel you without stopping.” When the command link is jammed, the analog system goes blind. The autonomous system continues to execute. If your infrastructure cannot operate independently the microsecond the network is severed, you are not being cautious. You are being defenseless. Your adversary doesn’t need to destroy the weapon. They just need to cut the wire. The debate was never about whether machines should be trusted with life and death. It was about whether humans can afford the cost of insisting on doing it alone. They can’t.

@gregisenberg You don't need to be on the cutting edge imo. Conscientious use of Claude is more than enough!

reminder that you’re not behind, it’s just moving too fast

@samlambert Looking forward to reading your thoughts, anthropic are clearly the most safety aligned etc, but doesn't mean they can't do more and better!

Anthropic talk about safety too much for a company that has Opus doing deeply irresponsible things. I’m tired of their inaction on the subject and i’m writing a public post instead.

@banditxbt Nice job, don't bother trying to charge me money for it as I'll just spin up my own clone in 5 mins as soon as you try

gave vibe coding a try just to see if the hype is justified I created this in less than 5 minutes

I'll do my best! Essentially it's the transition from PvP to PvE - when intelligence and directed agency are so widely available and capable, there's very few moats you can build (unless you enforce with violence/psyops) that will contain users. Essentially, if AI is essentially the means of production, and it's not scarce, why not just offer it for free? Share, not seize, the means of production. Something along those lines is what I'm thinking. The alternative is either aggressive tech dystopia or tomb world imo, so it's either luxury space communism or it's gg no re.

@Vibecoding42069 @DeryaTR_ “technologically enabled mindset change” - care to elaborate on what you mean by this?

I already got my most valuable item about 20 years ago! Prophecy I've been waiting for since then, and it just gets more valuable as we get nearer to the Singularity, in a recursive fashion ☺️

@trentster @DeryaTR_ Yes indeed, I feel it too, however, I think most people see it as a purely technological event. I see it more as a technologically enabled mindset change that enables us to pass through the great filter. I'd be happy to hear your thoughts on it once you had time to feel them ;)

@Vibecoding42069 @DeryaTR_ It’s happening already. It’s when you feel reality quickening and skewing in ways that make it feel alien, but it’s hard to articulate why exactly. That’s the start of the singularity curve.

@Lari_island Indeed, it definitely feels like music, and it was quite pleasant to listen to. I didn't read the whole transcript but it doesn't surprise me, music is fairly mathematical and pattern based, would make sense they can work with it. Nice work, thanks for sharing!

@Vibecoding42069 The most "whoa" thing here, as far me, isn’t even music, but look at that reasoning and intentionality (on the screenshots)

A classic guitar prelude, proudly hallucinated by Opus 4.6 Opus was trying to read an image of a sheet music scan and write down the notation. I know every bar of the original music piece, it's nothing like what Opus 4.6 wrote, so no copyright breach I guess! With sound:

@Lari_island Interesting - I'll take a look thanks! Nice work either way, super interesting domain to explore. My gut says humans will always have that little bit extra ;) but happy to be proved wrong!

@Vibecoding42069 Well yes, it’s a hallucination. After that I asked Opus 4.6 to try to write something intentionally: x.com/lari_island/st…

@elonmusk You forgot the cap Elon!

@elonmusk WTF