David Shapiro (L/0)

Amateurs. Y'all don't know how to manage your hyperfocus.

I've been watching too much Clarkson's Farm and writing historical fiction so I've just adopted a very dry, sarcastic, British attitude with my AIs and it seems to really work. Who knew?

Claude Code is unusable compared to Grok Build. I gave CC very explicit instructions about which folder and file to go into and it couldn't even get that right. When you've got the smartest (allegedly) AI on the planet that can't figure out "drill down into this directory and find X file" means, I don't care. It's a defective product. Grok Build set the expectation for instruction following for me, and admittedly I'm late to the party and Grok Build is the first terminal based agent I've used, but by comparison, Claude Code is basically VIM with extra steps, which completely defeats the point of "agentic" anything. At that point just copy-paste your code into the browser, it's easier.

@DaveShapi @RobertJBye that looks like claude web chat. Have you tried it in Claude Code instead? Or through terminal/CLI

Would love to see what you all work on next week!

@Awesome_O_AI @RobertJBye Yes, that's why I call it a defective product. I don't care if it's a hamster on a wheel giving itself a heart attack. That doesn't do me any good.

@DaveShapi @RobertJBye Burning tokens without results is incredibly frustrating.

@MatthewBerman The ouroboros of our time

@signulll @cortadoedison "No girls allowed in the autism app!"

I love being alive

First, no. Second, it burns tokens on stupidly simple tasks that 5.6 chews through and keeps going. So I can't even verify whether or not it can succeed because it's a vastly inferior product. I once used 40% of my session quota on a single prompt and it failed. Meanwhile, I have a long running thread with 5.6 with no issues across 20 iterations. No quota issues, no failures. So, my experience says that Fable is basically unusable for coding tasks through their own interface.

@DaveShapi Whoa whoa hold your horses Fable 5 still mogs gpt 5.6 in like.. 90% of tasks, chill out vro

I actually don't even care anymore. It's too censored, condescending, and burns tokens. Take it back I don't even use it. You got completely mogged by Sol.

@flowersslop The obvious answer is because it is theater and performance meant for a very particular, small audience. Virtue signalling through and through

If this were even a fraction as serious as they claim, the quirky self-branding would be grotesquely out of place. You wouldnt perform like this around genocide or human rights abuses. So if this is about something even bigger and even more serious, why doesnt it look like it?

Automation has not, in fact, been "net job creating" in aggregate. Automation has spent decades hollowing out manufacturing to the tune of millions of jobs lost, even when you correct for offshoring. The US manufacturing sector has about a third less employees than it did at its peak, but is three times more productive. Jobs that never came back. To all the developers and engineers who think that AI will not do the same thing to your industry: good luck with that.

has anyone figured out how to get ChatGPT to stop writing linkedin slop?

@BenPielstick I mean it's pretty cool but anthropic and OpenAI have explicitly been working towards it for more than a year

@DaveShapi How significant is Sol post training Luna autonomously?

You best start believing in fast takeoff... You're living in it

Dear Twitter, You have completely made Community Notes pointless. It seems like the notes community only uses it comment on the accuracy of memes and fandom characters. Not, you know, politics. The real world. So, if you're not going to enforce any sort of quality control, you should just disable community notes entirely. If it's not being used to fact-check reality, I do not give a fuck. If people just use it to argue over fandom accuracy, they should be kicked from the community. It's pointless.

Yeah as a former infrastructure engineer it's still wild to me that anyone lets this unproven technology have root access to critical systems. Do people not know what a virtual machine is?

@RammerRichard @ns123abc Right but that also just opens up... probably cause to search all his private shit too. Which is worse for him lol

@DaveShapi @ns123abc He told her to switch to a private Messenger AFTER all the incriminating shit was done. Too funny

>be Chang Liu >senior system electrical engineer at Apple >8 years working on iphone >january 2026: leave Apple to join OpenAI >apple asks for laptop back >ignore them >lmao it’s my laptop now >within HOURS of leaving >message Yu-Ting “Alyssa” Peng, friend at Apple: Liu: “I still have another computer” >uses it to access Apple secret info >within weeks, use HER Apple work laptop >february 9: try Apple’s network storage >cloud repo of confidential engineering files >authentication bug. still works! >message Peng: “LOL, I found out I can access the [network storage], so funny” Peng: “I’m ready” >while developing hardware for OpenAI >download DOZENS of confidential files >including a thousand-plus-page compilation of technical files >including MLB (main logic board) manufacturing + testing presentations >send Peng links to Apple’s proprietary folders >point her to specific project data >coach her how to copy files “to avoid trouble with the security team” >tell her which confidential Apple materials to study before her OpenAI interview >warn her another guy “fumbled” Tang Tan’s questions about a secret Apple project >“download some info” for her to review >tell her: switch to LINE Messenger so nobody sees this >she gets the OpenAI offer, leaves Apple April 16 >meanwhile every message was left on APPLE-ISSUED WORK LAPTOPS >july 10: Apple Inc. v. Chang Liu >named first. before OpenAI. before Tang Tan LOL so funny

Wait did it literally do the joke?

@flowersslop True. The "let me organize your emails" is basically cowork or computer use.

I dont think its even an outrageous take anymore to say ChatGPT will reach Samantha/OS1 level capabilities by next year, maybe even by the end of this year. Nothing she does in the first half of Her seems impossible or far off. It just needs to become more unified and proactive.

@HgcMatt Right but that's not even a good analogy. Machines are not children, and that metaphor smuggles in so much that we cannot assume about machines, which is also not even emerging. It's a poor model, and should not be taken seriously.

@DaveShapi Hinton used the example of adults taking power from a group of 5 year olds. If you grant that ASI will be able to think and plot circles around us then I don't see any escape other than augmentation.

All Effective Altruism, LW, and associated conversations have an infinite regress problem, and they always converge on anti-human conclusions. Here's how it works, with their own words. 1) Amanda Askell talks about "rational resentment" i.e. "the future machines might hate us for how we treat them today, so we should preemptively grant concessions and be nice" which is Roko's Basilisk dressed up in philosophical mumbo jumbo bullshit. 2) More and more conversation about "worthy successors" and "passing the torch" either because humans are "unworthy" or "will be judged" or because it's just the right thing to do. Sometimes it's also dressed up as "the only way to survive is by passing the torch" (surrendering sovereignty preemptively). 3) That we may "owe moral obligations" to the things we are building (which does not follow, unless you explicitly and deliberately build in a way that it can suffer... which is itself wrong but even if you make something that can pretend to suffer does not make it so). So, now when you look at the lunacy that is AI 2027 and their equally delusional follow-up AI 2040, it all regresses to "The AI will kill us, cause us to kill ourselves, and the only path to salvation is prostrating ourselves before the AI, or otherwise it's a suicide pact." This is all purely eschatological language. And these people are actively power seeking.