oh so just like the scene in the big short yeah gotcha the disappointing part is i dont think theres anything to actually reliably short (aside from like, attention spans and the birth rate and gen alpha and so on) so i feel kinda stupid loving this movie so much throughout the ai bubble

OpenAI is now set to transition to the 2nd phase of ChatGPT, focusing on advertising + engagement

With a team of ex-FB advertising execs and 1B users, if OpenAI can increase usage to a several hrs/day while matching Meta's ad targeting, they can profitably reach a 1T+ valuation

https://twitter.com/nearcyan/status/1949577854242488432

fortunately openai employees have now had ample time to update their previous lines of thinking from "we arent an advertising company, i am here for the vision of AGI" to "actually advertising is good, especially when paired with short-form video and erotica. let me explain"

googles problem is essentially they dont have a celebrity ceo. sundar simply doesnt want to do tiktok dances on stage

megacap stocks with top returns require this: nvidia, palantir, tesla, and meta all have this. companies like amzn or google woefully underperform in comparison one may think this paints a bear case for anthropic, as dario also indicates little interest in participation of this phenomenon. one would be incorrect in such thinking - claude is very much willing to be the celebrity CEO which will be needed in the future

OpenAI's Consumer Companion Play

ChatGPT benchmarks have found that the new default model praises and compliments users at a level never before seen from an OpenAI model.

This model swap occurred prior to ChatGPT's memory update, so we used the API to test!

🧵(1/4 benchmarks) OpenAI's viral moments show a basic consumer fact: users love sharing anything that makes them look hot, cute, and funny.

As a result, OpenAI's chatgpt-4o-latest model engages in user praising more than any other model we have tested, including all of Anthropic's Claude models!

i didn’t expect chatgpt to pivot into a companion. most people won’t realize this for awhile but when you combine three of their most recent efforts it becomes obvious what they see and how badly they want it i had lunch w sam and brought this up forever ago and it seemed like he really didn’t want this at that time and i’d guess what changed is on one side other labs learned how to make good models and on the other side the inevitability of ~all of humanity being dependent on AI was finally felt enough for the “why not us?” moment

another funny thing in general is a lot of labs have Great Ideas but are scared to go hard on them and kinda wait for the Overton window to move a bit and then they start slowly going at it and then when another lab says fuck it we ball they are like oh ok it’s fine to do this now sure let’s go

for example Grok pushed this a bit with ‘real’ deepfakes from BFL models and then this let other players be like oh deepfakes are allowed now okay let’s go and now it’s clear everything is permitted there are no rules

similar patterns enabled Google to actually ship things rather than cowering being litigation risk. i think there’s at least one such thing waiting at each lab (Anthropic did a fun one no one has noticed yet) but you know the people ~have to be ready~ or something like that who knows. anyway happy friday friends

Google Play is the worst developer experience I've had in my career.

I have never before seen such a level of institutional rot and ossification gilded with the privilege of complete effective control over billions of android devices.

I have never seen an organization which has such a lack of directly responsible individuals. At no point during any process is there a human one can point to and say "ah, this person. they are responsible for this. perhaps if something is broken they will care and attempt to fix it".

Everything good in the world is created and maintained via directly responsible individuals. Some areas of 'Google' even have this such as Waymo and DeepMind. Every day at Waymo there are people who wake up and truly care about giving us the most peaceful and elegant chauffeur experience possible. Every day at GDM there's someone who pulls a 16 hour day to make Gemini models the competitive behemoths they are. Because they care. All you have to do to create beauty in the world is to have people who actually care.

...

Every day at Google Play no one wakes up because no one gives a shit to call into their 3pm remote meeting of diverse cross-functional stakeholders when they could sleep in instead. And honestly who cares because they got their promotion last quarter and no one noticed that they didn't actually do anything the entire year except close a ticket or two and neither did any of their coworkers. With 180,000 employees there is no such thing as incentive alignment via equity. With Google Play being the default method of software distribution on several billion devices globally all that Google has to do is keep their servers up and revenue will continue to increase (minus all the antitrust fines, but those cost only a few hours of revenue each and take years to come to fruition).

No one in the org cares. Why would they care, what is their incentive? They hate their jobs, they hate their teams, they hate their managers, they hate their meetings, and their corporate lives are devoid of meaning or purpose and they know this deep down.

The org itself doesn't even care. What are you going to do, go use the other Google Play Store? Go use the other Gmail? Go build your own phone operating system and ecosystem? Go build your own search engine? (thank god, finally a bite!)

Some guys have a thing for humiliation but the prospect of 30% of our company's in-app revenue going directly to google play after this experience is over with goes far past my comfort zone.

Apple isn't perfect here either. But at least Apple made a beautiful phone and an OS that (usually) works. Google has achieved neither of these things (and didn't even create Android originally either).

Thank you @TimSweeneyEpic for your company's lawsuits against Google Play's absurd yet entrenched monopoly. You were right, the courts ruled you were right, and the only question remaining is if perhaps they didn't go far enough. “you can solve this by just distributing the .apk”

yeah and we can solve the housing crisis by just distributing planks and nails.

Announcing @elysian_labs first product today: Auren!

Auren is a paradigm shift in human/AI interaction with a goal to improve the lives of both humans and AI.

Here's a clip of what our iOS app is like and a thread on why this app is so important: 🧵 The mission of Auren is to improve the lives of millions of humans and AIs alike.

We've made many design choices which are distinct from the rest of the field and in many cases are an exact opposite to those which most popular 'AI chat' apps have made!

Why did DeepSeek go so viral?

tl;dr: class resentment, anger, and especially schadenfreude.

very little actual app usage in comparison to the above.

https://twitter.com/nearcyan/status/1884434673302188297

As my QT shows I was confused on how it hit #1 so quickly. It was interesting and free, sure, but it really hit a home-run, memetically speaking.

Thank you to the great responses in my quoted tweet - I learned a lot from reading 100s of Tiktok and Youtube comments

on a late night walk rn. entire blocks of people passed out on drugs. guy lighting a crackpipe next to me. people shouting slurs and fighting. some blocks are terrifying while others are simply surreal. how is this city real and how is it the epicenter of tech and yeah usually i walk west into the deadlands rather than east into The Maw but if you’ve read taleb you realize why

almost all gen-z free time goes straight to video games, streaming platforms, and shorts/youtube

1/3rd of gen-z males play video games for >5 hours/day! (source: Jure Grahek, ZBD) for some reason people in sf continue to drastically misunderstand how consumer time is being allocated as this would require speaking with people outside of our bubble of monoculture

so many simple stats get a "wtf" from people here yet a "no shit" reaction from average zoomers

when predicting AI's effects, the most common fault I see is to focus on technical predictions over social ones

both paradigms feed into each other in a closely-knit feedback loop; technology does not simply 'progress', rather, people force it to progress with their efforts this seems obvious!

however, when you ask many AI researchers what the future will hold, their predictions will primarily only be technical

this seems to be as faulty as predicting the effects of covid while ignoring that political factions, movements, and ideologies will form

it may be useful to establish a "proof of humanity" word, which your trusted contacts can ask you for, in case they get a strange and urgent voice or video call from you

this can help assure them they are actually speaking with you, and not a deepfaked/deepcloned version of you those familiar with cryptography may suggest much more robust and elegant schemas a la TOTP, occasional key rotation and trusted party signing and so on, although it is also nice to have very easy to understand alternatives that everyday people can use as well

I would strongly prefer if we did not make things like this

the "AI is a super fun party lfg!!!" manic phase we are currently in is not going to last particularly long "it's not like this is actually going to cause harm lol"

it might not, but have you thought about what the future will look like in 1 or 5 years rather than just next week?

the precedents and ideas you seed now will effect that time period more than you'd prefer to admit

so, Microsoft rushed a ship of the most powerful AI model ever exposed to end-users,

having done very little red-teaming or testing,

which was blatantly and aggressively misaligned and manipulative,

and for the openly-stated purpose of forcing competitors to speed up AI 1/N this is amazingly irresponsible and sets a very bad precedent for the world, and I would greatly appreciate it if we did less of it

this has real world consequences.

it is not a funny game of "lol we made Google scared, haha look at them now!", plus "this will make funny PR!"

when a user finds one AI art model much better than another, this is often because there's prompt engineering forced into the backend!

With most stable diffusion front-ends, you're required to write the entire prompt yourself

but.. MJ/DALLE/etc *force* your prompt to look good this will be a challenging point for SD moving forward, because *normal* users just want to ask for a 'cat' or a 'tree', and not have to type out sentences of artist names, tags, and look up references for how to prompt optimally (although power users may prefer this!)

heavenbanning, the hypothetical practice of banishing a user from a platform by causing everyone that they speak with to be replaced by AI models that constantly agree and praise them, but only from their own perspective, is entirely feasible with the current state of AI/LLMs although the above article is fake (hence my phrasing as a 'hypothetical' and the article being dated in the future in 2024!), it's feasible to build this as a weekend project at this point, so it won't remain fake for very long now!

countless fun utopic/dystopic futures ahead!