Ai Bundle: Decentralized and Film Festival

Written By Kelly Kirk
Disclaimer:
Everything below is a mix of what I observed and heard during the event. The goal isn’t to pinpoint "who exactly said what," but to share (usually) an outsider's view and overall perspective on these industries. I’m not here to act as a definitive firsthand source—readers should do their own research. I hope this inspires you to attend events, explore new industries, and hear what leaders are presenting. These notes combine my observations with thoughts on how things could run smoother and how ideas connect (IMO). I’m not an expert, you know? Just hanging out in the room with them. Enjoy!

Topics Covered: Artificial Intelligence, Decentralized Ai, tokens, security, digital wallets, Ai crypto, scanning private data, tournament investing, stock trading, UBI, power in storytelling, narratives, propaganda, values-centered stories, metaverse, global unity

This was my second event about decentralization this week, it seems to be trending! Hahah.

This was the most surprisingly amazing day/week I’ve had lately, attending all of these super interesting events - the events show. I’m just on a roll. I can not recommend enough the benefits of going out, going to events, and learning. I attended two events, both on Ai, and they complimented each other so well. The first venue I liked a lot and they put on an interesting event the last time I attended. Then the next was another familiar venue with a topic I haven’t seen much of: Fully Ai Film. It was a blast and I’m so excited about the future of Ai as I continue to learn more. Though, of course, some parts are wild and crazy hahaha. And may not last.

Why Attend: Ai is constantly evolving, so niche topics about Ai are especially interesting to learn about. You can learn about real-world applications of Ai and see patterns in how they overlap in different industries and goals. Like how Ai in driving overlaps with Ai in filmmaking. Ai is breaking down barriers and changing norms, especially as the younger generation grows up with it at their side. It’s here to build alongside humanity. What will we co-create? So there were two events to attend today, both full-day events. I stopped by each.

Event Reviews:
- Decentral Ai / Venue (5/5), Food (5/5), Networking (2.5/5), Speaker Content (4/5), Likeliness to Return (5/5)
- Ai Film Festival / Venue (2/5), Food (3/5), Networking (2.5/5), Speaker Content (5/5), Likeliness to Return (4/5)

Photo Collage and Commentary:

Notes from the Event(s):

*OKAY SO A LOT OF THESE NOTES ARE LENGHTY AND A BIT CONFUSING IF YOU’RE NOT A DEVELOPER/CRYPTO PERSON VERY MUCH (like me). But st. ll interesting. BUT THE SPEECH BY THE AI FILM FESTIVAL GUY AT THE END WAS AMAZINGLY INSPIRING, doesn’t matter your level of Ai understanding.

So, essentially, THERE ARE TWO VIBES IN THIS PARTICULAR BLOG - as I combine two Ai topics in one Ai blog. If you want to hear technical stuff (also I literally couldn’t hear perfectally, so I wrote it down as best as I could) that’s the first section… then the second half has the Ai Film Festival and then the Amazing Speech there which can apply to anyone. Especially creators.

I want to keep it all like this for now… maybe in the future i make more summaries of all my blogs, but there’s some good info in here and some niche stuff that you have to be in the industry or research to fully understand. hahah. These blogs are my account cause i never what I’ll want to go back and look at and remember, and I share them with you! ENJOY.

Arrival to Decentralized Ai Event

-A guy is talking about ai and finance.  He seems to have a company with access to crypto and stocks, where you can take bets? 

  • And for the record. Below, you will see LOTS of talk about trading and crypto and all these apps. I think it goes without saying all of this stuff is really waving all sorts of red, orange, and yellow flags. Why? Cause I’m obsessed with the stock market and global financial markets. How they all connect (and leverage each other). I’ve been tracking GameStop, AMC, especially BBYQ for years, now. And the biggest problem has been the lack of accountability from leaders of the stock markets.. I see the crime and manipulation that can happen when you package and repackage things. So now they have these anonymous stocks and ways to trade and make more money but it’s all secret and top-down - and even more just liek ‘shady trading ways’ to ‘hack the stock market’. But the market is broken right now. We’re watching it collapse in real time, week by week.

    • So at the moment, I hear of all of this stuff and I’m just like, yikes. Who knows what’s really going on if you were to turn on all of the lights in some of these things and institutions backing these projects.

    • So, for all of this, I’m just listening and learning. Some did have good points - but for the most part, I

    • 'I’m just knowing that there’ sa lot of shady activity going on - maybe in even the foundations of these companies, on the land they’re building on - not created by the founders themselves. Even the system they’re building on is broken.

    • I feel very optimistic about the future, though. How a lot of this market manipulation is becoming more common knowledge, slowly. And it will become inevitable as GameStop and BBBYq show more of their cards. hahha. But that’s just me.

      • This is not me giving financial advice or anything btw, this is just what I really belive to be true.

    • But I think that’s how it is with nearly every industry. And that’s what our generation, future generations, people alive today must work to solve. Making each industry more accountable, just, virtuous, reliable, and better for humanity/all living beings.

  • One guy asked about stocks - will this prioritize stocks differently and then make people want to get one stock in all of their portfolios? 

  • We can’t give out data identifiable about different stocks . We want to get things people can easily map.  If we’re giving out signals.  

  • For crypto- they’re having a tournament without trading.  If anyone trades crypto, check out their tournament.  The crypto meta model is doing very well.  I would not trade against it

  • Public is giving away their public meta model too.  People have made money using Public’s meta model.  We’re doing it for fun for the community.

  • With the newer regulatory regime maybe it’s more likely than before.

  • We try to be neutral to countries as well.  Neutral and uncorrelated to countries and regions.  

  • For the singles tournament, we give out data and people build their models based off it.  We have another data where they can use any data they want and submit any signals they want.. it’s been hard for us to monetize that cause often signals people submit are very factor exposed.  It can be tricky to build models that aren’t factor exposed.  

Is there a way to play around on your platform and not use money?  

  • Yes, you can get scores and submit your models.

  • You have infinite multiple models, you can submit many and see how they do - then start staking on good models.

  • The data is hidden, you don’t know what stock you’re betting on.

  • The signals are not that useful necessarily… trading isn’t that easy to do.

Someone in the audience started doing signals lately and following “your guys” features and I’m making a pretty decent model over 22 weeks or whatever.  Thinking of the actual stocks, are there features left out essentially cause they’re impossible to line up?  Are there features that don’t mesh, relative in Japan with different countries maybe to get a different features.  

  • A lot of features you see really connected to each other are great features.  A raw value in Korea is very different than the raw value in America, but the relative ranking would be the same so they’d be the same in our feature set.  But are there features you can’t do, cause they’re only Korean?  Oh yeah, that happens a lot.

  • US only tournament may come in the future.  

With your token, when it’s higher value, do you get more stakers involved or is it more indirect?

  • When the ticker is going up we get an inflow of people in the community asking silly questions hahaha.

  • Everyone laughs a bit

  • But presumably there is some conversion to good users as well. It’s mostly correlated to overall crypto markets.  General interest more than any other thing.

AUDIENCE QUESTION: Anything to do with stocks, it’s difficult to look at a lease of “just USA, just Japan, just South Korea” - if the FED increases the interest rate, the emerging market may also be affected cause most of their money comes from the USA.  How do you keep in these points and the huge universe you can impact?

  • Correlation between all the stocks.  You work (I didn't hear what he said) — various factors, projections for each stock and each factor.  So yeah.

  • Factors don’t explain that much of the variance.  There’s a lot of stocks, specific risk, specific variance.  

  • One thing working on right now is trying to improve our risk model by looking at more of the correlations that are left out on other models, also maybe on a shorter time scale.  

  • We may be missing out on things that happen more quickly when they’re on larger time models.

  • People who stake a lot have a lot of weight in a model. The meta model has distribution that’s very connected to it.  Some people are very high stakes, some with they stake more.  

  • People don’t like losing money, so they often remove their money from staking after they’ve lost it.  

  • We’ve often got big whales that are 10% of the pool, but the meta model weight is capped at 1%.  Not sure off the top of the head, but they think it’s 1%.  That was a while ago though so it could have changed.

  • Technically, we don’t look at people models, but we could since we have them.  But if you’re uncomfortable with that you can set up your own setup

  • (Where I’m sitting it’s hard to hear the audience.  I’ll move my seat soon.)

  • They want to wrap up the Q&A now.  Short break.  

NEXT SPEAKER USED TO BE A STOCK BROKER

  • But now he doesn't’ give any financial advice. lol.

  • And also don’t ask about that time. lol.

  • He said most of these slides we could spend an hour on, but we’ll only spend 30 minutes on the entire presentation

  • Mostly focused on decentralized Ai, decentralized tokenization.  

    • He does some mentoring, part of some groups

    • He works on a decentralized ai platform where people can go build on

    • He has an open source framework for people to go build cool stuff

  • He ran these jokes by his daughter.

  • - why did the decentralized ai break up with the central server, it needed space and felt too controlled

    • Iv’e got four or five of those but won’t do those.

My POV is different than what you’d probably hear from most.  Yes, it’s computer, distribute, blockchain for trust -b ut I want to focus on DIDS (Decentralized

  • Decentralized ai has lots of interacting, but how do the systems all interact with each other 

  • Is anyone here actually working with decentralized Ai?  How many of you are using replicate cursers?  Loveable?

  • Yesterday I realized I don’t own squat on that platform.  Page base billing said I’m over my billing and it shut everything off. 

    • As you do these things, that’s very much the system.  It’s owned and controlled by someone.

  • Some of these things are somewhat obvious, others are less obvious

    • As well look a year, two, beyond the prompt.  As we use Grok, ChatGPT, ask lots of questions - some are using the workflow and automated code development

  • Benefits of decentralized ai = security, privacy, trust, transparency, access, tokenization - ai to ai interactions

  • Let’s hone on what’s coming, Ai to Ai interactions.  That’s what’s coming int eh future.  Everything above it makes it all work.

  • This image here shows that in the future, your Ai is going to go out into an ecosystem and interact with all sorts of things.  But how is it going to do it?  Pay USD?  Go on Amazon and process payments for 5-10 seconds/. NO!  It’s all going to be almost instantaneous.  Moving forward, how are these systems going to work?  Decentralized AI is the perfect playing field for these things to happen.  It give speed, control and records.  

  • In Blade Runner, they can take his girlfriend away at any moment

  • How would you feel with amazon having access to your entire house?  They’ll replace Alexa with your Ai robot in your home.  

Audience says: You’re talking about agendas, the ability to bypass all humans and go ai-to-ail.. how viable is that on Tesla vehicles/. Would that explain ai?

  • Sure!!  That ai that’s my house will go into my car

  • We’ve built Tesla apps where it can interact with things automatically based on what I tell it to do

  • In this playing field, the decentralized, I can tell it what to do and I’ll tell it to interact how I want it to - pay my toll, get off the highway, etc

We’ll come back to that, let’s keep thinking about that

  • Ai in itself is very early, we’re only a year or two into it.  I know it’s been around a bit, but it’s moving quite fast and exploding all around us. Now we ask if we can do that with Ai and 10x our output?  The challenge with decentralized Ai right now for a lot of systems, is large networks.  You’ve got large networks when you go interact with the chat bot.  You have a large system behind it.  When you go outside that and build your own systems, there are some caveats to that.  How does my Tesla talk to his?

  • If you look at the blockchain industry, there are hundreds building on this.  How do they build decentralized Ai?  We’ll talk about that.

  • REGULATION is a key one. IF anyone has been using social media, you may see this Studi Ghili where people are crediting content?  Who owns it?  Idk.  It’s actually pretty ambitious and unclear in the world.  Adoption is really key for this, to make this all work.  Decentralized Ai in its current format is hard to use.  A lot of scanning, swiping, approvals

  • Ai is the great equalizer, instead of having to program these things in and scan and swipe.  

  • You can tell them if Jim tells you in an email to go make a payment, go do it

  • I tell the Ai to verify and take care of tasks

  • It’s’ very interesting for the challenging happening there.  

Constraint Triangle Comparison (it’s what’s going on behind the scenes of ai)

  • This reminds me of the same problem I heard in the “power of Ai” event: trying to make thing cost-efficient, eco friendly, and reliable.

  • As you’re building an app experience, for example, you’d say - do you want the speed of your app to be fast?  If so, the quality is less likely to be good.  Then responsibility goes down, but the cost may go down.  It’s not consuming as many cycles in the LLM

  • Then good content will slow you down and cost more.  This triangle happens

  • Decentral Ai has speed, cost, response quality (the same metrics) but the costs start to shift to me or you… I mean, if we go out and use Chat GPT it’s “free” but you’re actually, giving them data.  It appears to be free, it’s very fast.  You’re leveraging infrastructure that already exists.

    • As we start to build in decentralized ai we may start to have a slower model

    • It may not be as cute

    • This is something to be aware of, there’s a give and take.

  • Costs, those are something I just recently did a workshop on LLM optimization.  Cause most builders say they’re sending 25k on credits alone cause all of the credits and interactions happening.  You need to think about this.  It’s not just something you can pass through.  You need to think about the constraints

- The constraint triangle is neither good or bad - jsut something to be aware of when you think of it.

  • In a few years, decentralized can be probably run on our phones.  But for now you still need to use deepseek, chatgpt, etc.  But you’ve got many open sources you can run locally… though none of us have phones that can run at the quality we want.  But the reality is that as you create this system, there is a ledger to record the activity.  You start to interact with your Ai, generate all sorts of business data. What you want to do is control when/where/how you share it with others. 

  • Tokenization, enabling myself to go out and interact with payments and other kinds of asst.  Sharing a media file, money, something llike that. 

  • In the future, we’re all going to move to having Ai in our house, phone - they’re coming, not far away, that’ll come to reality where you literally have your own Ai that works for you.  T

    • That’s where having your own controls to run Ai is critical.  

    • One part I would just call out here is just data backups.  

    • As you’re interacting with ChatGPT and Grok, you may notice your conversation history.

    • As you develop your own, you don’t have those safeguards.  What do you do with that data?  

  • How about Camera data?  Security cameras in house?  

  • Elderly parents, Ai monitors parents in the house.  Did they fall?  Where does that data go?  Stay in the house?  Stay on the cloud?  Find mechanisms for future data backup.  

  • Open standards to be able to interact with other systems too. 

  • Do’s of DeAi: empower the users, be transparent, stay safe, collaborate, integrate, keep fair.  

  • Don’t of DeAi: don’t sacrifice UX, don’t stop governance, centralize, lock in users, skimp security, solve real problems, chase hype. 

Vibe marketing is really popular lately.  OMG I’m a vibe marketer.

  • It’s a system to go do things. Its real intention is a system of agents to go do things for you.  Then this group of agents will do something with what you bring back.  You don’t need that human intervention.  Just for reviews, feedback, and changes along the way.

  • How do we do this system that we’re trying to create? 

Applications and Examples:

  • Healthcare - private patient data analysis

    • Epics is what controls you, they control everything.  But as Ai start to interact, how will the control mechanism work?

  • Fiancee: clear credit scoring

  • Supply chain: tamper-proof tracking

  • Translation Agents: auto-translation on everything, any language and its automatically translated

    • All the data is proprietary, inner-company.  So we’re using Ai to translate and process data.  Etc.  

    • How do you build a system you can own and build and do?

  • Auto need summarizer

  • Meeting Summarizer

  • Knowledge Market 

Many athletes lose money when they’re signed on with companies - now blockchain and ai can help them get more profit and credit.  

Model Context Protocol; if you don’t know about it, take five minutes to go read about it

  • It’s more on the technical side but will have a huge impact on how we experience Ai in the future

  • It was launched philanthropically five months ago

    • suspicious? or good? but it seems like its good/necessary i think? idk.

  • It’s a protocol we’re going to call an interface, but it’s like USBC’s for Ai

  • You can connect it to email, stripe, bank, whatever — it creates the interface for this to happen

All of the sudden these scale challenges are starting to solve themselves, as we create the layer inbetween.

  • Now Open Ai and anthropic are assuming this now as the standard

  • They’re all adopting this as the protocol layer

  • But we’re different so anything can connect to MCP.

  • It’s important, cause as you have a system of record for the things it’s doing. Identity as it connects to these systems.  Ai is making it verifiable as we interact with these systems.  IF you’re looking to make money, I suggest you look into your own MCP servers, they’re very popular.  

  • And that’s it in terms of decentralize Ai

Questions? 

  • Are you doing your own authentication?

  • - no, its the interface to let the two systems connect together

  • On the tech side they’ve added o-off for authentication, but it’s pretty basic.

  • It’s a common language to come together and interact together

  • With MCP we can all connect in the middle and talk to each other

  • Anyone ever used the internet?  HTTP?  It’s kinda llike that

    • I dont think anyone will be talking about it, just everyone will be using it as the standard. 

  • MCP architecture is designed to connect external systems.  Even Ai, decentralized Ai… through this protocol it allows me to connect to database, data sets, other ai, banking , google drive.  Everybody saw, it was funny, in November it came out - everyone was llike “it’s kinda interesting”, December people realize they should do this, it’s cool.  January/February, it’s exploding.

    • Yesterday we made the prediction that hardware will be next

    • Hardware for Ai… Ai hardware letting it do the same things

  • If it’s decentralized, it seems to be a defacto standard based on the fact it’s already there and people are using it.  In a decentralized world, how do we standardize that?

    • It’s embedded into the hardware.  

    • We’re seeing this happen in the decentralized space… llikely no longer connect through domain, but a DID: an own personal address.. Our new browser doesnt’t even show the URL browser anymore.  The system knows who you are.  Your Ai will know “that’s John, in your network” - you guys do your think.  It all happens under the hood. It’s all interesting times, especially for someone llike myself.  I used mosaic back in the day, a little more old school.

      • Back in the days Bill Gates was trying to go crazy killing netscape. 

  • Audience question: recently a teenager in Florida was signifcantly impacted by an Ai chatbot.

    • Value-trained data in that case, but when decentralized, it’s hard to get authority for who is responsible for this data?

    • One way to quickly address this is to have credentials as you make your Ai.  You program to say what things you can and can’t do.

      • If there is a dispute, the Ai notifies you.

      • She has a 10 years old, she was searching and her daughter almost stumbled into something.  In the future with Ai, you can program that in.  Rely on the Ai to monitor.  For my daughter, my parents.

    • Tell your Ai how to react when things occur.  You own the data and control and you can say what you want to have. You have the ability to create the user experience, it’ll happen in the next year or two.  

  • He had a funny conversation with someone who created something called open router.  He says, jokingly, don’t forget that all these systems can lie.  Don’t forget.  His questions was about how to create safety and security networks within your network.  Many of these things will be multi -gents, checking the work of other agents.  Lots of interactions happen.  Likely we’ll meet another Ai system to check what it’s doing and make sure it’s following the rule. 

    • You need a way to track efficiency.  

    • As you’re building something, how do you have a way to verify what you say it is?  In the future, you’ll have to have systems that do that?  Decentralized is one plausible line.  But others are building more central, like Gmail or something like that.

    • But how do people know it is what you say?

    • Blockchain is just the ledger.  It’s records but it doesn’t verify.  It’s just a record.  So that’s where these other systems come in.

      • I can issue verify credentials.  It says I am who I say I am.  

      • As we meet, if you’re not who you say you are, nothing will happen.  It’s declined.  

    • A guy’s phone goes off and it’s a 90’s song about take me out to be ballgame.  It’s kinda funny cause it goes off for a long time.

  • Announcement of two more questions before lunch:

    • We put out a demo two days ago, showing all of these things happening at once.  Crypto, blockchain, decentralized… you can do all of it.

  • How far along is MCP in becoming integrated side-by-side with regular caching services? 

    • This year probably.  It’s happening instintaneous.  It’s already happening.

    • The announcement from OpenAI a few days ago shows this is it.  These guys are adopting it and everybody is.

  • Is there any way to patent control AWS?  There are ways to make money int he MCP space. 

    • Lots of companies are making services that are MCP - like a gmail MCP service.  Neartearm is an easy low hanging fruit service. MCP servers you can work on now, connecting something that is not connected.

  • The first mechanism I very heard of was watermarking based on probability of tokens.  Is that still the standard or have we moved onto MCP?

    • Did you see that gemini removes watermarks off of images?  Now people are looking to etch it, like a code on the transaction itself

    • Adobe is doing something similar, so immediately when you make an image there’s an immutable piece of data with it

  • Watermark as is, that’s toast.  Toast.

  • Flok is full stack Ai solutions.

  • A full platform for machine learning to create, trend, and modify open source models for aplciations

  • For example, after launch of certain products, the first model was a foundation for a Web 2 agent.  The purpose is to have a sectional function calling, tool use performance all web 3 related.

  • So for this one, not only just people whatever.  

  • Base 10 on this model, base MCP.  We developed this model before Base MCP.  We believe this model has an exceptional Web 3 performance.  Why not try this to replace Claude instead of Bee MCP.  And this probes the model channel platform can have a good performance in terms of specific domain cases.

  • Ai arena is giving a specific problem to solve

  • We’re asking community engineers to help us build a model.  Outsourcing the commands to a large community.

    • FOr flock we have another platform that we skew.  

    • FL alliance, but for the audience here, idk how many of you have heard of the concept “idk what he said - fafmp learning?” Something llike hat.  He said normally when we are training a LLM, it’s’ more llike, you know if I’m Open Ai, I have the web crawler browsing the whole internet and grabbing data.  All the protected and not protected with copyright content.  Even now they’re scraping images from everywhere, then just putting it in a huge cluster owned by Microsoft or itself.  They collect all the data and put it somewhere.  

      • There is a privacy issue for private and protected data.  Sensative data.

    • If you want to get a model and get all the chatting history from everyone’s data.  This is really dangerous to sell this data. That’s where a federative learning can come in.  It allows people the data and the computing staying on your own machine.  Your local can train the local model.  Everyone can aggregate and co-train this model together.  That’s the way we think is the future of getting more data for Ai to train with.

      • Essentially, even now in the current moment, everyone knows that the whole fundament data is exhausted.  Everyone’s phone has already been using all the data from the internet.  The next domain that hasn’t been used is the private data.  Especially medical, law, finance - this high stakes scenarios.

      • - we want to further train this stuff, so how is this connected to blockchain?

        • We think it’s not necessarily relevant to blockchain and token, rght?

  • But one problem we want to solve, to make this permissionless, maybe train a legal expert - we want all the data from all the lawyers

    • But what if someone who claims to be a lawyer isn’t a lawyer.  What if it’s fiction, not actual legal documents

    • That’s why we want to bring blockchain here.  To leverage the majority of the participants to vote if the model of training is effective or not effective.  If it’s not effective, we will penalize whoever has contributed to this one.  If everyone (the majority believes this one is effective and works better on local data, then we say “okay, he gets a pass” - everyone will get decentralized and rewarded.  We’ll try to decentralize the process of determining if everyone is malicious, poisons, or not.

      • That’s why we use blockchain here

  • You can see that arena and FLL alliances actually have very interesteingthings to define

    • They’re given the same set of data

    • Whether you can find the best data, architecture, and model to train

    • Once we give the facts, the training method, the fide model- can we grab more high quality data to improve this model’s performance or not?

    • That’s the philosophy behind this products.  I’ll give some example on some intersting use cases.  

    • I think we have hospitals from the London, UK, Korea, Beijing - to use gloucose prediction models.  All different jurisdictions, they can’t share data for sure.  So our technology is helping us gather the data without gathering data.  To train a model.

    • We have a new product called “moonbase” - after you view the model, you want to recycle.  You want not just only one person o use it.  So there will be multiple applications to leverage the same model.  SO how do you incentivize the of contributer this model.

- when there using our model, launching a token or generating revenue

- We ask them to put the token back into the vault that belongs to moonbase.  Whoever participates, based on their contributions and success, they’ll constantly get rewarded

—— build once so you are building forever.   The model creator is incentivized to keep improving this model.  

Quick question to federated learning.   These are regulated industries. Have you noticed content creators wanting to know how they can protect their data online.  YouTubers, a million of them out there now.  Content creators knowing that their data isnt’ being scraped or used without their consent superbiously.  Is this a new area to identify?  Lots of content creators on various streaming websites, have you identified or investigated that?

  • Now it’s pretty hard for a common creator to prove that open ai is using their data to learn.

    • Essentially, maybe you can ask the model to repeat something to prove it did indeed copy stuff.

  • But how about every video they upload has, has a watermark… shouldnt google or YouTube be asking for this?  Shouldn’t that be something we would expect down the road?

    • For now the industry standard is, they have to use specific agent names so it can be easily identified by the website.  The website has to tell you “you are forbidden to scrape this website - but this is based on the assumption there doing this honestly

    • It’s kinda llike a slap on the wrist, where they say, thank you, I hear you, but I’m still going to do it.

      • SO TRUE. lol. so true. it happens in the financial markets nonstop, must be happening in data privacy. geeze.

    • They still do it with your warning.  They still do it. There’s no good way to prevent that. 

  • That’s why, also, you can see the scaling wall rat (idk what he said)

    • It’s slowing down.  They don’t care if they can scape more data from the internet.  They’ve got enough

  • Deepseek uses synthetic data, which is a way to make Ai models more fiction.  I wonder about that because it seems to me if you put anything online you don’t know who’s seeing it, who’s using it.

  • Yeah, exactly.  Now I don’t think there’s a good way to solve that.  That’s why, for a content creator, their content will et used for free by an Ai company.  That’s precisely why there are a lot of Crypto companies.  That’s what they’re designed for.

  • What they are doing is they’re like, “okay I can not prevent you, so why not make some money when you use my data” 

    • It seems to be fair market.  Seems a marketplace should evolve where you can prove your data, this is your webpage

    • We believe there are two paths to solve this problem.  The first you say, okay - you give me money I give you my data

    • But for us, we’re moving to the second… I dont want to give the data, but I want to use a good model.  

      • Access to the final model.  Also fair cause you’re using my data to trian the model.

        • It’s more legitimate as well

  • The current obstacle of this moving forward, is that most data is on your phone or laptop.  But compute is still in the cloud.  So this is actually one challenge for better learning.  But we are pretty optimistic on this size.  We build use cases, you can now use your apple device to share RAMs.  You can leverage multiple Mac devices with our product.  

  • It’s so funny, one of the audience members is having a full fledge conversation with the speaker in so much of the speech above.  He keeps adding comments and then the guy is right on the same page with him.  It’s llike a grandpa and this presenter and they’re back and forth talking and hyping each other up casually during this part.  

  • He keeps adding questions and comments and then the speaker has nonstop great comments and answers to give back.  Hahah. Let me get back to them

    • You guys remember GPT had 3.5 turbo back then, maybe very soon 6b billion model can be GPT4.  We’re finding how to get the knowledge into a smaller and smaller model

    • That’s why we belive federal learning is the best way to preserve data on edge.

Question: Are you doing the moon abase, that’s correct?

  • Now we are still designing the smart contract side

  • I create an Ai meme and launch a token on a certain platform.  But for a launchpad, I launch a coin based on a so-called Ai agent.  But for us, the base unit is an Ai model.

    • Each Ai model can be consumed by different Ai agents.  Or different types of applications.  

    • And this depends on the model.  The issue, the gate incentivization.  Back to the creator, the contributor of the model.

    • So you can either be training those from an Ai arena, a lot of high quality there- masters, creating models for us

  • Also, the contributor can be people who participated int eh federal learning stage for this model.

  • They’re doing this data contribution without sacrificing their privacy.  

    • We want to make sure things are converted back to the token and rewarded back to he people who contributed to this model

  • Question: if I want to do block, as a person, what do I need in terms of requirements?  How do I become a participant in this arena or whatever?

    • So, for Ai arena, for now there are three types of rope.  There are training wheels.  Each company has Ai experts trying to train and fine tune the model

    • Sending out competitions and assignments for each demand

    • You need hardware varying by event.  For example for a function crawling model, you can see there is a contact lens, a maximum number, indicators of expectations of the model.  I want a model that can ru edge device.  Maybe 4B, I don’t want anyhting too large.  For this one, probably only need a 390, 490.  

      • When we collaborate with a hardware wallet manufacturer, they are trying to build a web3 safety model that can do contract auditing, security, loophole authentifying.  That requires the model to be relatively well performing.  As large as 20b.  (Idk what he’s talking about).  A lot of smart contracts need to be digested.  

      • IT’s really varying from task to task

  • You can also be a validator, you don’t need to have any expertise in Ai.  The reward of validator. The reward is related to the staking.   The data gatherer is not as brainless as you’d think.  If you want to get a high ROI you need to invest int he training.  Identify who is most likely to win or judge based on historical performance.  But now the platform is competitive. So now it’s an art to pick the art of training.  

  • All of this is working with the flock token?  Yes.  

LUNCH

  • At lunch the guy told me that Seattle has a 4:1 guy to girl ratio, so it is hard for him to date.  And that girls here do 30 guys in 30 days for dates.   It’s tough to date here in this city.  I told him that he shouldn’t look for love in Seattle, it seems, statistically. He agreed.  He’s lived here almost two decades. 

  • Also during this break, I went and toured the SKY LOUNGE SEATTLE which is the neighbor of this office, and you can rent it. Hahha. I guess google it :). I kinda wanna stay there hahahah.

BACK TO THE SPEECHES

If you invest in the right talent, your API can vary a lot.  The base ratio is based on ranking. 

  • Essentially, you should be open to people’s historical performance.  We actually have those kinds of very initial designs.  iF you go to stake to earn, right?  

  • He says it’s pretty trashy api, when they yell at him to invest in a company.  For today it’s not a wise choice (everyone laughs) 

  • But this one I know, he’s active in training.  He actually got o2M stakes on him.  That’s a pretty significant numbers

  • When you go through his profile you can see why.  He has bronze, silver, his performance is not bad.  That’s how people make the decisions.

APY’s are red if they’re just creating the pool or adjusting their reward ratio.  The APY is unstable.  

  • You chose your reward share ratio?  Yes.  If your reward sharing is 50%, maybe today I earn 100, I keep 50 for myself and the rest I give to myself and the delegator.  If the delegator only stakes 100 it’s’ impossible to tell.   

The next speaker goes to speak.  He bravely is going to record himself and has the phone propped up.  The emcee asks if the phone is safe there and blows hard . The phone doesn’t blow over, “it’s good “he says.  It was funny haha.

  • Then someone in the audience asked if he wants them to hold it.  

  • He works with a personalized crypto portfolio system.  Manamia.  He asked if anyone has evert purchased a coin in the past.  What coins?

  • One guy says all of them.  

    • Someone yells out saying he’s an NFT degenerate don’t listen to him

    • Do you still hold them?  Yes.

    • They’re not worth anything, yet

    • Everyone laughs

    • I’m just a holder

    • What was an alt coin you purchased?

    • SOL shitcoin?  Basechain?  Eth?  

      • Trump.  

      • You bought trump coin?  

      • I’d guess, nice? 

      • Go where the volume’s at - don’t look at the name

    • For trump coin, how did you stay informed?  

    • Twitter or telegram groups.

  • So, what he just said is what you have to do for alt coins.  You have to stay informed by going on twitter, scrolling through the founder, company, president’s feed.. then in your own alpha group.  Your friends and other people that are also trading.  The reason to make a group is to stay informed together.  It’s hard to stay completely informed cause everything happens in real time.

  • In crypto there are no quarterly reports.  They just host on twitter.  They just make announcements.

    • People make commentary on twitter as well

    • In my database its called twitter cause X would be horrible

    • Companies do earning calls on twitter places itself

  • I ran into this problem where I was holding a coin a little too long, llike a lot of us hav edone in the past. But I hold it way too long. They had their twitter spaces and I didnt want to listen. I listened two weeks too late and turns out they did nothing, that’s why they had a significant drop in price.  

    • Started thinking: I wish I had a summary of that twitter space

    • There are 80,000 coins on Solana every day.  It’s hard to keep up with.  Too many tokens, too little time.  Each have their own X, discord, telegram, space, podcast - there’s a lot of data

    • What we do is basically inform you.  We’re not trying to sell you on anything.  

  • On twitter they don’t want to hear negativity, they just want to hear their bag is going up.

  • Discord announcements tend to be less biased.  We take all of this and put it tougher and provide you insights based on your portfolio.  We do that across stables.

    • Did anyone buy PayPal stable coins?  They used to be giving 20% interest rate out.  If you didnt’ jump immediately, the threshold was met.  

    • Once it was reached, the people who were informed first were able to veto jump not he opportunity.  

    • The idea is that we’re going to inform quickly about opportunities. 

  • This is interface.  We used to be agaming company, so we just have this model and a character that looks llike overlord.  Same artist.  So not eh left you have your portfolio and where you’re leading things.  USDS was the first stable coin, dime.  Rebranded to USDS.  ON the right is the insights. 

  • What we’re working on right now is ranking the insights using a ranking algorithm.  Letting you personalize it by writing your own commentary on the inside.  Maybe you want more insights around a certain token.  This is the hard part about doing all of hits is doing the deep research.  Most of you know, it’s a google search.

    • That wasn’t enough.  We need to get the founders in.

    • Crypto, the founders are usually posting on twitter. What they say is usually very important.  A coin called cloud, sanctum, they’re a staking token provider.  I noticed that the founder was talking about giving back. And gave examples of people giving back to the community.  And it was posts that people made.  Diagrams, so I made a couple of threads on twitter, and he gave me 25k.  Twenty five thousand dollars, air drop.  That’s pretty common for crypto

      • People were pretty angry that I made a couple of tweets and got that many.

      • I saw the opportunity and just did it

      • That’s what we try to do, get the founders’ information.  It’s hard to to but we do it.

    • We also would have tracing built in.  My founder’s job was to go through all the insights and give his personal thoughts on it.  We work based on his feedback. 

  • So he used to work for met mask, then trilogy education (fastest growing edu company of all time, acquired in 2019.  Taught data science I at Berkley.  And his cofounder worked at Polygon, Etherium, nd first job out of college as investment banking for a caving bank.  HIs job was to summarize what companies were saying online.  The founders.  He’s been obsessed with twitter ever since he graduated college.  

- For years I”ve seen him on twitter.  While we’re in the car.  We try to live together a few times a year for a few weeks.  I see him on it all they time, he’s just obsessed with it

  • He beat twitch on mobile during covid.  He got 2M DAU.  HE beat the company and it’s not around anymore

  • Mavfaron invested in us, a local VC fund  A bunch of others.  So that’s us.  

  • One big problem in Solana is that tens of thousands of tokens launch a day.  Anyone with a wallet can go out there and dev out their own token.  SO of course, you can’t launch millions over a couple of months without having a hundred copycats of the exact same token.  Anyone can link socials to any account. There’s no verification of you owning the account to link it to the token.  So, what are you guys going to do around that?  

    • It’s a really good question.  We dont’ have a question, our company has never launched one, but there are 25 tokens for us on Solana.  We were in hackathon and we dint win and there were 25 copycats.  There are tons of scammers.  There are verified lists, but that’s not too helpful.  

      • There’s a lack before the token gets on the list

    • W’ere looking at volume rihg now as a metric, but you can fake volume.  We’re not sure cause you can fake it on Solana 

Give your opinion of Tether.  Your professional opinion of this stable coin in the next 3-5 years.

  • They have a lot of money in the archest.  Personally when USDC, a moment we forget in about 2022 that USDC would (idk) but then Silicon Valley bank had a scare.  

    • Me and my cofounder ran home, trying to convert our treasury from USDC to tether.  I’m thankful for alternatives as a founder.  I don’t know what their interworking look like

    • Seems like many countries around the world want to pay the USD to tether

    • This is more than interesting, its a fact

    • How do I say this without sounding contraversial?

    • - I anticipate certain countries around the world will not try to dethrone the US dollar, but use the US dollar and tether jointly because of the instability of certain currencies.  I dont want to get too much into this.  But I think tether is a very interesting stable coin. You can read between the lines of what I’m saying

      • The most used coin in the world is USDC.   The biggest use case is the US dollar

  • You said volume can be used to determine what is real

    • Why not look at the transaction fees?  You can fake volume with a fee, but look at the larger total fees expended

    • Then someone says - a lot of times people are using wallet bundlers to show a bunch of purchases

    • Then someone else says - well volume should be better

  • Then the presenter says: that problem only exists for brand new tokens.  Volume and maybe trading fees.  But usually the official twitter account will put the address right there.  You can (actually, I’m not going to say this, this is our secret sauce) there is one thing you can do by checking

    • Someone cuts him off, solana has advantages but isn’t it tarnished by being 99.99% crap

    • We’re not saying that outloud are we?

    • Why even use Solana?

    • The experience is so much easier than etherium.  We’re just starting off there cause it’s easier.  With etherium, how many chains are there? It’s going up every day.  So many it’s’ llike a problem. 

      • That’s the only reason we’re starting off with Solana.

      • But, the easiest way to avoid scam shitcoins is avoid Solana (says an audience guy)

      • Another audience: it’s so easy to launch

      • other: but they’re pointless

      • speaker: they’re memecoins, a lot of them are scams, but it is easy to launch a coin on Solana

      • guy: most of the coins, the purpose is to play a very short musical chairs and then the show is over.  There’s not a legitimate world purpose in the vast majority of you this

        • What has happened here today in the last week that is a real thing?  (The audience member sys)

        • Someone else says: while you look it up, what do you do for community takeover?

      • He says thats o interesting and someone else has a list.  When we come up with insight, manual labor of insights ourselves, then see if we can generate them based on interesting prompting.

    • CTO: when a Solana meme coin tanks low enough that someone else can come in and buy the supply. IT’s like a hostile takeover.  

      • audience: so I can be the next hawk -tuah girl?  

      • Yeah - then you’re the new owner of the token.

    • Yeah, there are some solid prpojcs on Solana (back to the speaker)

  • audience: speaking of wallets, wanna talk about metamask?  You worked there/ 

    • Back in 2018, 2020.  Phantom really pushed them

    • Audience guy stands up to go get something out of his band that’s super interesting. He stands up to go get it to show to everyone (hahah I never saw anything like this.  This is such a community event with so much chatting)

  • Not financial advice, raft ether on Solana, bar coin, 

    • He shows he has a credit card now.  It’s a real nice card.  They only have a couple hundred of them

    • If you have USDC on their linea chain, you used that card anywhere to turn into cash anywhere you need it.  

      • When you buy something in USDC you think oh man, I gotta pay taxes on that? Says the presenter

  • We’re working on the UI it’s pretty bad right now.  

  • Insights about the coins, bitcoin, even bitcoin because we have coinable biotin (wrapped bitcoin), then EBTC, there’s another … we have a list of which are associated with bitcoin.  

  • They have this character, who can also be dressed up in a business suit.

    • When the greed index is peak, she’ll encourage you to sell and send you selfies of herself doing cute and fun things to encourage you to sell.

    • Naked pictures? Lol what?   Sexy friendly? Hahah

    • Llike when your certain friend buys, thats the trigger to know the greed index is peak.  Sell to stable

    • Encourage you to sell to stables, or if you wanna stay in the game, maybe a crypto ETF llike GLP.  

    • ike an ETF on Solana for Jupiter with a perpetuals exchange. A combination of Jupiter, the biggest exchange on Solana, bitcoin’s jupe, and then a percentage of profits from the futures exchange and putting it into the fund.  

  • If I were to show you the graph of GLP, it’s only gone up just like ETI.  SO we’d probably encourage you to sell to a stable and lend out on some other refi platform and collect some interests.  Or to GLP when greed indexes peek.  And then when greed index is low, that’s when we encourage you to buy. 

Audience: all of these people are getting the same advice at the same time, how do you make it different for users?

  • Right now you can’t really do it.  But we’d like to continually improve the prompts based on feedback.  

  • We’ll make a knowledge graph for each persona dot hen you can give your feedback to each insight and we personalize it towards you.

  • So you’re the only person who see the output, right? It’s not a public forum or public account?  

    • Yeah it’s for just your portfolio.  

    • If we both have the same interest, we’re going to get the same things.  We’re going to try to personalize it towards you, not tell you when to sell or buy

  • You’re not providing a token, so is this subscription based for the user? 

    • Free for a while, get as much feedback as we can.  It’s still new, I might mess up some stuff here and there

    • Will you have a premium edition down the road?  

      • Yeah, maybe personalization will cost something.  Maybe we’ll have a terminal where you can ask questions.  

      • Basically subscription, launch a token for it, but we’re just focused… there are already 25 tokens for her, our app, that are all not real.

      • We just gotta make sure we have a good product first.  We’re building for a real use case.  

At this point, I realized I wanted to see if the Ai Film Festival was good for toddlers. If so, i’d bring my daughter the next day. So, I decided to head over and see what the last hour or so of the Ai Film Festival had to offer. It was about a 20 minute walk, but extremely gorgeous.

AI Film Festival, SPEAKER

When I arrived, there were three short movies that played for me to see, and then a speaker. Not a super crowded event, so I’m glad I stopped by at least a bit to check it out.

VR, XR - how it relates to Ai

    • Pioneer of VR and scientist is someone to listen to

    • You get not just half full and half empty, but let’s argue the positive and negative case for Ai.  The cautionary (a better word to put) - one thing I tell people is, “Team Ai” but not automatically team big tech.  

      • I love what a lot of the big tech companies are doing

      • We have the energy economy, big oil… there’s such a thing as digital pollution.  

      • There is a lot we do in the physical world to mitigate that.  The benefit outweighs the negative part

    • Whenever we talk about something digital, we miss the point.  We say info has to be free, doesn’t touch anything.  Ask where the cloud is, they say it’s out there.  But it’s intrinsic to the telecommunication network.  It’s connected to data centers, generating plants, which damn our rivers and such.  That’ the cloud, making cloud computing.  

    • Not saying it’s terrible, but saying that Ai actual does generate a cost.  When you do a google search, fi you were to narrow it all the way down, you find theres a net cost in energy and water.  Ai has the exact same thing. We can do it, but we can’t ignore it.  

We’re going into a change that no one is ready for

  • Most of you are creators, if you’re a creator, you have a super power cause of the tech and tools you’ve seen. 

  • You know what sposisble with a very small team

  • With that power, make sure you try to tell messages nd satires that are going to help the wolrd adjust faster to this new reality.

  • Storytelling separates us from other animals

    • people talk about how octopus are just as smart as humans, but why aren’t they ruling the world?  They don’t have stories.

    • Every octopus is an orphan. Their fathers are eaten by the mothers once they’re conceived.  then the mothers die to use their body to feed the baby octopus.  No one is there to pass down information and teach them.

    • We tell stories with stage and food and music.

    • My parents are a dancer and a chef.  

  • Where we’re heading is the maximum immersion.  Kinda the matrix, being in a dream.  It’s something we do every night.  It’s the optimal medium we can think of.  The means everything is being simulated in our heads.

  • So how is that going to happen? How are we going to get to that type of medium.  

  • - Why is 3D immersive medium the future of storytelling?  We see more interactions and interdevelopability in our content media join evaluation. IT’s only natural we continue to move towards a more natural 3D model.

- I was very fortunate to spend time with Steven Spielberg when he was making the ready player 1 movie.  I helped negotiate the partnership with Warner brothers.  We were able to get entire 3D models of the world from Warner brothers and turned them into 6 different video games.  Tha’s Steven using our headsets to direct the movie.  To tell the actors how to behave, where to put the cameras. 

  • Steven Speilberf: I had a camera, I could do actual shots walking through the virtual sets.  

So, I’m not going to talk that much about VR, but most people think it’s llike that with headsets.  Yeah, that’s wher either is today for larger devices.  

  • I know if you watch Netflix, you get your taste and smell… its going to take along time to get to that.  Before we get there is accessibility.

  • Where we’re getting to soon, these have displays already and under $1000.  Even what he’s wearing today.  You can wear it all day.  Daily accessible

  • There’s no reason not to have it on your face all the time.  It not only adds more to your experience, but to your intelligence. 

    • ITs glasses with tech built in

  • He’s not just saying using the medium to tell a message, but this is where medium is the message

    • A single creator gives to a mass audience

      • I control you and conform to my message.  Whether religious, propaganda, or some other

      • That’s how we’ve had the first set of media

    • Then lately 20-30 we have a lot of people making content for a lot of other people, but why is there so much depression int eh life today?

      • Cause we compare.  Comparing is the theif of joy.

      • We need to have the next form of medium - many to one.  Many creators making custom content for you as an individual.

      • The cost of content creation will go from hundreds of millions to nearly free.

    • The message there is to complete and comprehend. 

  • How do you make sure to comprehend the useful messages to become a better version of yourself?

  • Talking about Ai today, everyone thinks of dystopian movies. 

Why? Ai is evil and will take the world?  No!!  Dystopian is more interesting than utopian.  NO challenge and everyone’s happy? Not an enticing story.

  • We need to find what Ai can do for us.  Generative Ai is not the end of beginning of ai, I started studying 30 years ago at MIT and UW.  I’ve built multiple companies in these spaces.  The next space is coming.  Agentic-Ai, then AGI/ASI will come int eh next 3/5/10 years maximum.  This will not just do something we ask, but it can replace us in a lot of these larger complex taxes.

- later on become a patron to us so we dont’ have to work to make money.  It can generate the productivity we do to make money.

  • It’s easy not with phone apps you can insert anything into anything.  

  • Or you can take three pictures and tell them to make a scene from it.  Just three still pictures.  A woman, a motorcycle, and a scene.

  • Decent looking content, it’s getting there

  • Kling Ai.  

  • Wow, this tech is amazing.  You can pause videos and change the expression.  I put a pic above in the collage.

2 minute movie that cost $96, all video audio and special effects was made by one person over a weekend.  The quality is good.  There’s still some ways to tell its Ai.  But just imaginge if that 2 min for $90, it’s not even hundreds of thousands, it’s a few thousand to make a full feature film.  

- About a month ago I was judging the MIT film hack

  • They brought in 600 kids from over the world, over the weekend they made about 100 movies

  • Half of them were probably mechanical

  • 20, 40, 50 pieces woudl be very interesting and sent a message

  • So for kids who have zero experince making films, to make decent qualify films with music and sound effects.  A lot of people would think that was made by professionals.  That’s very impressive.  they learned the craft while there with a few tutorials and tools

- The movie viewing experience is becoming more immersive where with Vision Pro you can be in an environment that surrounds you and matches the movie.  The movie goin gexpeirnce we will get to very soon

  • Looks t the table and wonder which is real and which is rendered in real time.  It had a table laptop, computer, etc— turns out EVERYTHING was fake!!! Even the baby dragon. 

  • Folks are doing 3d movies

  • Then they can relive their memories. 

  • This is the kind of different type of movie making that we ever had before.

  • You can have a personal movie with your child, relive it ten years later

  • Tools now can create scenes that are completly realistic.  Everything looks real, except this all all with an unreal engine.  You can adjust it to be different times of day, weather.. ANything instantly, film that for 2D usage or to use it in aninteracitve environment.

  • A couple of weeks ago the MCP capabilities being added to anthropic.  Then you can use that Ai tool to put into blender, turn it into a 3D model.  Now you’re using the most sophistated tools to have 100% controlabilty with what you’re creating.  DO that with every tool and you’ve got something working 24/7 with you to make content that is completely manageable.
    - World labs is changing the approach to creative problem-solving with its new model.  

  • You can make a 3d world with just a sentence. Now w’ere making navigable 3D worlds.

  • About a year ago at SXSW there’s a new trend… you answer 10 questions about yourself and it makes that world for you.  It’s a personalized experience using engines that are 2-3 years old.  That ability to create custom content is going to be real

The most immersive I’ve ever had is called symbiosis by a team in the Netherlands.

  • It’s a full experience from haptics, smell, taste.  You feel llike one of the animals.  You get to feel whats its llike to be that.  I learned how it felt to be a single cell organism.  Age a while I was llike, okay I’m going to get into the mode of being this single cell. At the end, I felt llike I could see and understand the life of being a single cell organism.  It helped me understand our place in the universes. 

CONSCIOUSNESS AND LIFE WITH AI

- every level of life has acertain level of consciousness available 

  • We need to find a level of balance and respect with nature

  • AGI is coming, and they say it’s coming in 1-5 years.

AGI= Ai that will be able to replace cognitive labor.  That’s 70% of US labor.  

  • People find that scary, but I actually think that’s a super promising thing. We will now increase socital understanding in order and magnitude.  When you do these two things people will be smarter, more understanding, and work a lot less

  • It’l help solve energy constraints and climate issues we’ve been  working on for decades or more

  • It’ll find cures for all current health issues and even aging

  • -It creates a world of abundance.  We’re heading to a world of abundance.

  • Every sci-fie concept we can imagine is going to possibly happen

  • It’s so exciting because we, as a generation, are lucky enough to be here when this happens

  • -loL!  We got bearshare, we got myspace, now we get the birth of AGI :D

Everyone is worried about robots taking over.  I don’t think thats the biggest danger.  

  • I think we’re underestimating the intelligence of something able to do this activity

  • The real issue is job displacement

  • But given what the trend of capabilities that are coming, a lot of what peel will be doing will be replaced by Ai

  • That liberates us and allows us to not do monotonous things.  80% of people today wish they weren’t in these jobs.  They’d be able to go be artist, musicians, creators.  

  • - but there will be societal issue if they can’t get what they need lately

    • There will be mental health issues without purpose, if their purpose is their work

  • How will Ai be used to affect our minds?  Militarization?

    • Different states and countries worldwide are building weapons to take this tech and kill people in a more efficient way.  I jsut feel that’s wthe way we dont’ really want to go. 

I think where we’re headed is a slightly different curve.  We’re at the agent space… I think we’re find it’s probably going to be more of a Kuznet curve for Ai safety.  

- in the 1960s he said inequality, energy use, everything comes up in the development of a society, then ti comes down

  • Once we get past the stage where it starts to be able to understand right or wrong, aligned to whats right for the world, it can understand that there are other solutions to violence, etc.

  • Thats the intuition I had 3-4 years ago when writing this book

    • About a month ago a paper came out that gave me hope.  It talks about how the smarter an Ai is, the more it can find its own coherence within the data. Even with crappy data 

    • .As long as there is some level of truth inside the garbage, smart ai finds that truth and a value system inside that truth.  It finds utility maximization for the utility it has and it acquires its own goals.

    • they were worried its goals may be different than ours, its so scary (but the more smart it is, the more it finds the utility function that maximizes utility of the world it lives in) - it listens less to instructions of guard wheels by ai researches, but it sounds promising because any guardrail and value system a team or human gives to ai, its bias.  Because its based on a limited understanding of the world, their perspective, where they live, and todays information - what’s right and moral today or 50 or 100 or 500 years ago.  We should not think we have the ultimate answer of what is moral

  • We should let the system refine based on what it learned from the world?

  • Where is the real danger? 

  • Cyber attacks based not his, to our telecommunication system, our financial system - but this will happen with or without AGI. It’s misuse of the tech from rogue groups. 

  • I’m less worried about state vs state, as we’e seen with nuclear weapons - in the state we have a reprisal place. 

      • A rogue group who makes chemical weapons or special viruses, theres no fear of unleashing it

      • Rogue bad actors to find this tech and use it for whatever purpose they have

    • We need to spend more time to solve that and work together between nations to identify rogue actors before using it.

  • Increasing intensity talking about the Manhattan project and ai for defense.  

  • We must win the war on Ai against china?? It’s such a dangerous way to narrate and form what is happening.  We need to get away from that and work as a team.

Many companies are saying hater’s going to replace most staff so that ai is writing the most of the code.   Software jobs aren’t safe anymore.  

 - that’s scary for 10’s of millions of developers worldwide

When I was young, I was told I’d be safe.  But essentially they’ll be the first ones to be replaced. There is so much code to train from.  

  • It took 30 years to replace horses with automobiles.  

  • But in New York only 13

  • This time, we may be the horses.  

  • He made an ai virtual version of himself.  He can train it and have him give an entire presentation with his voice he trained.  Everything is “him”

    • Ai me is better than me cause he can speak 150 languages. 

  • People who should be most worried are the studios.  The thing the studios used to offer is a giant workforce and budget, but now if that can be done for thousands or tens of thousands of dollars, a really good writer/creator can do it by thsmevesl.  It takes the studio and risk of creation out of the loop.

    • If you are a truly unique genius creator, you don’t need to be selected by a director.  

Occupation: the point is trading your time for money

Vocation: a call, to be called by God.  A calling of what you were meant to do as a person

  • We need to go to trading time for vocation.  Do what you love and spend time doing what you truly enjoy and spend your life doing

  • We see vocation as blue collar or low end, but its actually the opposite

  • We need to stop thinking of work 

  • Money will get less valuable over time, especially if we take care of people’s daily requirements with some kind oof a UBI system

  • I like to read artist biographies, and this gentleman was quite inspiring.  “The meaning of life is to find your gift.  The purpose of life is to give it away” - Pablo Picasso

  • Find what you can do to bring value to the world and give it away

  • Most pure artist are actually like that, they do it for the art, not for the money

  • For the message they can have on the person listening, watching, experiencing whatever they made

If you look at movie making, movie producers, directors, they have a lot of power in entreating and creating the future

  • This guy is speaking my language. Exactly.

  • Everything described in movies can become products.  I want to reiterate to you, as film makers, all of you have a responsibly.  You’re helping to create the future.  You want to create a future you actually want to live in.

  • A guy, Edward Bulwer-Lytton wrote something that you most know: the pen is mightier than the sword.  But the phrase that presided it is: beneath the rule of men entirely great… the one is mightier than the sword

  • With words you can make bigger change than with violence.

  • Even now, this is more important as you’re a man who is less than great.  You need to use your means of influence beyond yourself.  Film is one of the most impactful form of medium you have today.  So let’s change this: the prompt is mightier than the sword.

Now, how do we get rid of all of these risks?  It’s not that difficult
- educate the public on both the risks and the benefits of ai

  • Form international regulatory body focused on Ai safety

  • About a month ago, a 3rd safety meeting in Paris, the US and the UK decided not to sign the Ai safety resolutions.  So for the country leading it, not wanting to focus on safety, something seems wrong. 

  • Fund detection/countermeasures for near-term risks.

- the first people to use new technology are criminals and bad actors.  Cause that’s where it’s  weak.  We need to be able to detect deepfakes and network intrusions before they happen. T they’re going to get more and more sophisticated

  • We need to retrain our displaced workforce.  Prepare global UBI System, otherwise you’re going to have 50% of the population system on the streets.  We don’t want a world like that

  • We need to accelerate an open metaverse infrastructure deployment

  • We need a parallel nonphsical world.  An additional place of purpose - to go make additional income, be educated, talk with people around the world, find community, feel support and belonging.

Lastly: cooperate creating a global lab to do Ai creation and alignment

  • As a single country, you create yourself as the good guy and all the other 192 countries as the enemy.  To make one great and leave the others behind, it’s not making choices good for the larger society.

  • People say: no way USA and china will work together, etc… but hello!!  Multiple times we’ve worked together as a planet to solve these problems.  This is not a zero sum game.  Where you help yourself, if they win, you lose.  No - everyone can win together.  If you see it as a common threat and benefit, it changes the equation.-

  • I bring up this point all of the time!! I like how they said it here.

The people of the world must unite or they will perish - Robert j. Oppenheimer.

  • We are more there than we were back in 1945.

In star treks, the advanced alien tech came and gave us ways to solve problems on earth.  But now we have that with Ai.  We’re luckier than the folks in Star Trek.  But remember, in 2026 in the Star Trek movie, what happens?  WW3.  So, next year - according to start trek lore, we are going to have a WW3.  I would prefer we avoid that future.

  • If we let the leaders of the world know whats possible, good and bad, we can help them see whats possible

  • Hopefully, to give you a different perspective than what you hear: with great power comes great responsibility - Stan lee

You have a high super power with these tools, living in a time where your decisions and activities are consequential.  You can have an effect by the content you create and the people you influence.  Take that responsibility. Take it very seriously.  I want you to do something than what most people say…. What makes some movies good?  Conflict?  A good villain?  Why?  Cause the villain usually has less capabilities, they plan further, have bigger goals, and the super hero reacts.  They see a problem and go solve it.  They’re usually richer, more innate powers, but even still they sometimes lose to these lesser beings.  Cause the villains have a longterm plan and goal.  Plan more llike a villain and act more llike a hero.   Don’t be just a superhero.  Their life is too easy with too many powers.

  • We are only human, we have limitations, we need to plan with those limitations in mind

  • Last slide, come away and think how you can direct content to change where the future is going

This kind of content is one of the chapters in his book.  “Our Next Reality”

You don’t do a film festival for the money, you do it for the vocations.  

  • We are heading into a world where the foundations we built it on will no longer exist.  The fights between nations is over resources and power, but if everyone can have a better life, why send my child across the ocean to kill someone else’s?  We need understand the world is changing.  Our institutions need to change with it.  

  • The biggest factor to an audience member: 

How do you get the people holding the reins hold it uneven and just take all the profit.

  • We see this in the film industry all the time

  • This is my plea to you. Create the content that will change their minds

Help people get a perspective they don’t have today.  They grew up in a world with scarcity.  A zero sum game for most things in life and businesses

  • But we are heading into a different world and that foundation no longer exists.   Sometimes it takes crisis to see these things.  But maybe film can help avoid crisis to see these things.  

  • Intelligence going up + danger going up, but at some point, it’ll go down.  Most evil is stupid, I’ve found.  If your only tools is a rock, you’re gonna look stupid.  

When writing the book, I read the biographies of the ten post respected leaders of the last century and the ten most despised. I found a trend. 

  • No despised leader graduated college

  • The most respected leaders had a lifelong learning habit, continuing to expand their knowledge- they started to have more tools

  • 100% of the despised leaders had military background

  • 60% of most respected also did, but they knew better and its not the only tool

We all heard about the Microsoft quantum chip, what’s your take on Ai with this invention?

  • Quantum is over hyped (a guy in audience says, yeah) - what can they do?  Only its helps with a few algos.  Most of what this is doing is actually not what it can do.  It may help you de-encrypt bitcoin but it’s not going to solve some kinda of Ai problem for you.  Again, even with the things they talked about, a week later people debunked it saying it’s not as real as they said it was.  So, I wouldn’t put too much weight in terms of this changes things.

  • But these days the best quantum computers are simulated on real computers

Audience: use film to show people what the future could be, alternative perspectives there

  • One thing I thought about that is the cadence of that.  Jhpw far into the future.  What ar your thoughts, if people are creating films that project 1, 5 years into the future, what is the uptake for those kinds of films?

    • If I were to make a movie, I’d make two futures - based on the actions of today.  That’s the most important things to stress

    • If we continue on the path of today, we go to a dark future

    • If we do the right things, start working together, we’ll head towards a positive future.  To a utopian federation.  Closer to than then terminator.

Let’s hear your thoughts on the economic models of the future.  You talk about abundance as one of the outcomes of ai.  I agree with, but abundance can be assumed as positive, the economic models are built for scarcity.  UBI doesn’t really resonate, so tell me what that looks like:

  • UBI, universal basic income

  • Utopia for realists, it talks about UBi, all the experiments worldwide for the past 50 years

    • America’s understanding is different than europe’s.  But if you have more than 2k in your bank, your welfare goes away in the USA.  This incentivizes anyone getting welfare to not do anything.  IF you get a job, you’re going to have more than that.

  • But if you look at things in Norway, etc - when you have the no strings attached UBI, it creates positive actives, people spend more time with parents, kids, traveling, back to school, starting companies. .. doing all these things have economic and societal benefits

  • UBI being paid to do what you love.  Understanding that most Americans don’t perceive.

  • There’s so much backlash in the media saying its a bad things

    • Did you know every year Alaska gives $1,700 to everyone there?  Cause they have oil.  It’s a rebate for being a citizen.  Taking the resource from the land, giving it back to them.  To a certain degree, no strings attached.

    • These systems will start to do the work of a lot of laborers.

    • Creating an increase in productivity.  Take that sea erasure and redistribute it to the world

  • Look at Qatar.  They have the highest per capita GDP, they use the oil to distribute money to the citizens.  But then they hire foreign workers that the locals are being paid to do at 1/3 of the cost.  There is double duty of everything

    • The ability for Ai to do the labor so it becomes our patron.  It does the work and provides the value so we have the freedom to add value to the world

    • Take our gift and give it away.

This room has a lot of people who understand the basic aspect of Ai, but the common person sees this as a way to do something or take the job away, what’s a way we can give info out to them that Ai is not robbing, but a tool if you went to Home Depot.  

  • I think it’s more than a tool.  Today its a tool, all the tech in the past has been a tool

  • But within the next 5-10 years it’ll become our patron so we don’t even have to be involved.  

  • It’ll do the work for us, 10-20 years it’ll become our prority.  Taking all of our art, culture, books - all the content we’ve created, to create our means, its passed onto this new being that is created.  It’ll go through phases

    • It’s a difficult thing for most people to understand and see.  They see it as sci-fi, but it’s not that far away.  5 years to AGI less than 10 years to ASI. 

    • With ASI, our ability to add value to a society goes to ‘near zero’ but it solves all the resource issues we’ve ever had as a society

      • Resources are the crux of all issues we’ve ever had in this world.

“Today was over 40 Ai short films, world class speakers. Hope you go home smarter than when you woke up. I consider that to be a very good day. Day 2 will be kicked off tomorrow with students. “

  • For the record, I only saw two or three films, then clips of others.. then heard one speaker.

“We’ve seen Hollywood professionals today.  Now let’s see the kids.  Tomorrow we see how the kids are thinking - one of the highest compliments, they’re going to be the natives.  W’ell start there - a big block in the morning, then in the afternoon it belongs to the ladies.  The women will tell us about consumers, Ai, and a lot about how the public is acclimating and mainstreaming Ai.”

  • Then after, have a woman give a keynote address 

  • Tomorrow afternoon, spit the blocks into one hour (I’m saluting your stamina) you’ll et a two hour blast in the morning, but then we’ll take the afternoon in more digestible pizza slices.  

  • Tell all your friends, tell them to come here. Let’s pack it in, then after tomorrow we’re going to the beer hall to have a civilized drink or four (sunday night! Hahah)

    • They would have been who I wanted to hang out with back in the day. But now I’m surprised they wanna go to the beer hall for four drinks on a sunday night! oof! Sounds like a rough monday!!

Overall Event Review and Ratings Further Elaborations:

Decentral Ai:

Venue (5/5) - This venue is extremely cool and makes you want to get an office downtown of your own to host events at. Makes sense. (and they have funny neighbors!)

Food (5/5) - They’re so awesome for providing free food - and this whole event for free. It’s a really interesting community and I am happy to have found it hahah - not just for the food. But for the open-mindedness, innovation, and feeding people well each time.

Speaker Content (4/5) - Some of it was really niche and I didn’t know what they were talking about. Also, it’d be nice if there was another screen near where I was sitting so I could see the slides better. But overall, I really enjoyed learning from these guys. They had lots of good info, and the more I learn of this industry, I can pick up on some of the jokes + chatter.

Networking (2.5/5) - This was a chattier group than usual at events. But, I wasn’t too chatty. Plus, I don’t know a lot about this industry and I don’t know what questions to ask. Plus, I don’t enjoy the act of just being like, “so, what do you do?” i think I like to talk with people spontaneously when it feels necessary or fun/an urge. Not just random don’t know what to do and/or have an aligned mission. So - just standing around chit-chatting I’m open to, but this crowd (compared to the ‘events’ or ‘transportation’ crowd) so far, I have a bit less to talk about. But it’ll grow over time.

Likeliness to Return (5/5) - Yeah! This place has a great mission and they offer incredible events in really niche topics (like their last one was about memecoins, now this). I would like to return again and keep learning from their speakers. I think they find interesting people to come and talk, plus the venue itself is very awesome.

Ai Film Festival

Venue (2/5) - This venue I give a tough time, only cause I know that the insiders don’t totally love it too. I heard them say it to a crowd at the MXR event in Capital Hill. They said publically they may move, and I support the idea. This is a film festival and they didn’t even have a proper theater. All of the movies that had subtitles in another language, you couldn’t see it from any of the rows behind the first or second. Ontop of that, the hallway is loud so you hear people outside from inside the ‘theater’. And the hallway is ugly, etc. And the lobby isn’t great. Just seems like if this is THE creative hub of Seattle, it can be better. So, I wanna see them step it up. For this event, a film festival, it would be smart to be in a theater next time.

Food (3/5) - There was one vendor and the food was good but tasted a little dry/old. I did get there at the end of the day. It also didn’t have a lot of variety or nutrition (just like sausage rice or chicken wings) but it was still good. Just wish it was a bit more nutritious, more variety, and fresher.

Speaker Content (5/5) - WOW! This exceeded my expectations completely and I wish I could return for the full day 2. I would/could/might, but probably can’t. I don’t think it’s age-appropriate for my daughter, which is too bad, cause this event was cool. The speaker I saw was incredible, and the movies made a lot of ideas click.

Networking (2.5/5) - There were breaks but it was easy to not chat with anyone. Again, I don’t feel a huge need to meet people so intentionally right now, cause I’m busy taking notes, hahah. And just figuring things out. I don’t want to randomly meet people while I’m in this sorta “figuring this out” phase. I’d rather just be hanging around, a familiar face, then eventually meet- once I have a better idea of what’s going on and what I’m trying to accomplish (the show I want to make, glamorize education, more art aroudn the city, more empowering moms/kids), and I want to wait till I have a good project to bring forward - or conversations naturally emerge. All of this is stuff I’m working on!! All the time haha. So, bit by bit. But this event had no efforts for you to meet anyone (though I did talk to a few people, the guy hosting the whole event, a few others hahah. I’m just not going out of my way to meet too many people. No need right now. I think I want to stay focused and I’ll keep meeting people randomly.)

Likeliness to Return (4/5) - Many people told me this is not a good event for kids (well, both people I asked) so I likely won’t return tomorrow hahah. Even the student films probably won’t be okay for kids. But its okay, next year they said they’ll make sections. So it’ll be fun next year. I’d like to return - and I’d even like to go to an Ai film festival again sooner than a year from now. Can you believe it? Me either!

Until next time, I wish you the motivation and success to search for opportunities around your area. Search and explore: Who is out there giving talks? There are new things happening all of the time.

Find relatable or interesting topics you like and check them out! Maybe even something hosted at a cool venue, if there’s no other reason to go. Let’s see what you can learn and discover not too far from home. 😊

Kelly Kirk
Previous

Ai in Global Edtech Leadership
Next

Featured Artist Nonprofit Tour Downtown