Will King

@posedscaredcity @n_sri_laasya @Experian @Plaid How funny would it be to actually do this and then the loan market gets entirely disrupted.

@n_sri_laasya @Experian @Plaid I can make one that approves loans to anyone, instantly

I built a multi-agent framework that pre-approves loan applications in under 5 minutes 🎉 1. Integrated 5 agents that make use of @Experian and @Plaid APIs for real-time credit and financial verification 2. The agents look up credit score, DTI, LTV, assets, property valuation and income to assess the risk profile of the application 3. Used AgentStack to build the system, and @AgentOpsAI for observability Full access in DMs.

@candlelovers12 @ask_aubry I'm not an apolitical person or a lying conservative and your stance is divorced from reality.

@ask_aubry Also beware of the "apolitical" people. They are almost always very strictly conservative , they just lie about it. If you're conservative or whatever just be honest.

We don't want to date ChatGPT.

@ExtremelyAvg @shouldhaveanima As a Will, I support this

@shouldhaveanima The face! So freaking adorable. I want to know the little tortoise's name. It looks like a William, but his friends call him Will the adventuresome.

@jacobcrxss @BreezeChai Gotcha yeah I feel you on that. I like the art you're making!

Is AI art truly art, or just machine output? What defines creativity?

@BreezeChai The creative potential of a blank AI canvas is equal to that of a blank fabric canvas. imo to be creative is to turn a novel idea from the minds eye into a tangible reality that others can experience

@Protogenes1 @BreezeChai If AI is a soulless product, why has its intro into my life enriched a drive within me to manifest something new? If AI art is so soul crushing, why do I now write, consume, & experience more poems, songs, & art than ever before? You are not your brush. Neither is your art.

@BreezeChai No it is no art, it is machine made. AI marketers are desperate to redefine creativity to get their soulless product accepted. Shun them, shun the evil ones

@yegordb To be clear: nothing about how your algorithm is designed supports any of the claims you are making. I didn't realize that @StanfordMBA had such lax standards and methodological rigour.

@yegordb I read your paper. You slapped together an "expert rating system" that you calibrated against a predictive model measuring "coding time" & "implementation time" Why is that a useful insight into coding productivity? You are a snake eating its own tail thinking it's feasting.

I’m at Stanford and I research software engineering productivity. We have data on the performance of >50k engineers from 100s of companies. Inspired by @deedydas, our research shows: ~9.5% of software engineers do virtually nothing: Ghost Engineers (0.1x-ers)

@arin_spanner @Hipst3rMusic @JudiciaIreview @eurydicelives Something few consider is how within group variation is usually much wider than between group variation. This kind of generalization focuses on between group & ignores within group variation. I am not amoral, & most men I know are not. Social bubbles exist & you're in one too.

@Hipst3rMusic @JudiciaIreview @eurydicelives Men are generally higher in ego/ narcissism and lower in empathy and I think increasing scarcity has made these traits more pronounced.

@truemeritocracy @OfficialLoganK @prasann_pandya You can!! @RewindAI

@OfficialLoganK @prasann_pandya If I could share my screen with GPT and talk with it live... total gamechanger!

Hot take: The GPT store is going to fail just like the plugins. I want to use products with specific UX and not use everything in chat interface.

If you can induce that state with code and #GPT can translate your thoughts into code then that could mean a world where you experience the #Metaverse in your mind may be closer than you or I ever dreamed. #generativeai #imaginationengine #howisthisreal

🤯 Oh look, there goes the future again hurtling towards our present. This is absolute insanity that we are so close to this! If you can induce a specific conscious state then you can induce another reality. #morpheus1

@moneywithkatie I live this truth daily. It is so easy to be generous and kind-hearted when literally every one of my needs are met. The amounts of money that change people's lives is shockingly low. And yet so much of my class hardly gives.

something I don’t think we talk about enough is how much wealth and free time can enable anyone to make decisions that look abnormally virtuous or honorable. sometimes what we attribute to good character is just money

@OfficialLoganK Couldn't agree more! Funny enough - after I started using ChatGPT, I found I was able to better communicate with my team and virtual assistant. If anything, 'prompt engineering' gives you instant feedback on the effectiveness of your communication.

@RobLynch99 @IanArawjo @ChatGPTapp @OpenAI @tszzl @emollick @voooooogel Did the effect size increase or did the p-value get smaller?

Interesting, I ran at N = 80 again just now and got p-value of 0.089 (two-tailed) but I did it on character count, not tokens (see my clarification to the original post). I ran it a few times over the weekend (making sure *not* to p-hack), and the effect definitely grew as samples did. I wonder why that would be impacted by tokenization though?

@ChatGPTapp @OpenAI @tszzl @emollick @voooooogel Wild result. gpt-4-turbo over the API produces (statistically significant) shorter completions when it "thinks" its December vs. when it thinks its May (as determined by the date in the system prompt). I took the same exact prompt over the API (a code completion task asking to implement a machine learning task without libraries). I created two system prompts, one that told the API it was May and another that it was December and then compared the distributions. For the May system prompt, mean = 4298 For the December system prompt, mean = 4086 N = 477 completions in each sample from May and December t-test p < 2.28e-07 To reproduce this you can just vary the date number in the system message. Would love to see if this reproduces for others.

@RobLynch99 @ChatGPTapp @OpenAI @tszzl @emollick @voooooogel It looks to me like you've p-hacked your way to this result. If memory serves me when the sample size gets around 500 for t-tests the power of the test increases to the point where it will post statistically significant differences for the smallest of effects

you can have 3rd parties generate and suggest possibilities, options. but you cannot outsource your taste, your discernment, your judgement

@vidhster This has the energy of keep your name out of my mouth and I'm here for it

2. Don't make questions like this required! Not everyone can answer them, and requiring them just increases the chances people will abandon the survey and/or that you'll get bad, meaningless data 3. If you must require them, keep jargony question names out of your error messages

Masterclass of a survey if you wanna learn what not to do 😵

@OpenAI can y'all please dismiss this undismissable notification so I can use ChatGPT on my phone again? You also shut down my Intercom feedback conversation before I could finish explaining, so here we are #qa #bug #error #killmenow