Podcast notes – Cosmos App Chain Thesis – Blockworks (Zaki Manian, Jack Zampolin)

Podcast: Blockworks Empire
Guests: Zaki Manian, Jack Zampolin

2015 in crypto – no one cared, maybe a bit of enterprise blockchain stuff
An email list – Google Group – crypto economic research, “Discord for old men” – Vitalik was on it, CEOs of early crypto cos
*Hosted one of most important events in crypto history – Vitalik didn’t take PoS seriously until he saw Jae’s Tendermint presentation, Nakamoto PoS

Ethan Buchman liked it, began to collab with Jae

Cosmos isn’t Ethereum killer, just as old
Jae’s mission was to make PoS be taken seriously

In 2016 – Jae and Ethan wanted to do public chain

Slashing was Vitalik’s idea – Jae helped solve it in a simpler way

Lots of what’s hot now were Cosmos early ideas – liquid staking, PoS

Introduced Atom in 2019
Early 2020 – Jae wasn’t easy to work with (eventually stepped down), Ethan wanted to quit, Zaki tried to keep things together

2019/2020 – once Cosmos went live, Binance adopted Cosmos SDK; didn’t need Tendermint anymore – Jack and Zaki quit, tired of working with Jae

Santi – Tendermint is battle tested consensus mechanism; Binance, Terra used it; Facebook re-wrote it in Rust for Novi / Libra

*Strong culture of testnets in Cosmos – influenced rest of crypto culture

Single smart contracting blockchain is a lot like a super computer – but there’s a max speed to it, can’t put all of the world’s apps on it

*App chain thesis – to meet demands of 8B people, need lots of chains that talk to each other, that’s Cosmos IBC
It’s about full stack optimization

Every app doesn’t need its own validator set – in shared security, Cosmos validators will validate your chain for you – use Cosmos toolkit, put up a proposal, and validators start running the software, earn rewards to do it

Cosmos believes in ship and iterate – get v1 out, get users, build on it
V1 – can support 10-20 chains – full validator set
V2 – will do partial security / partial validator set

Each chain has different incentive model to reward validators and early token holders

Cosmos top of everyone’s minds now – because of last 2-3 years in crypto, builders understand full-stack optimization and what’s needed
Iteration cycles are slower – 6-9 months for each

DYDX migrating over to Cosmos
Initially built on top of StarkX – similar to IBC; doing $1B of daily volume
But running centralized orderbook – tenuous regulatory position, not censorship resistant
Moving to Cosmos now – can integrate token, censorship resistance, Tendermint + IBC

Once app reaches product-market fit – can easily leave Cosmos Hub – don’t want to build barriers to prevent their departure, but hopefully shared security model is worth it to stay

Currently every Cosmos chain runs own validator set – doesn’t have true shared security yet

Every Cosmos Hub proposal gets 60-70K votes – very high participation

Cosmos is going to kill fat protocol thesis – popular meme
As Cosmos grows, app layer will accrue most value
Original whitepaper barely says anything about Atom token – in 12 days they’ll present Atom 2.0 tokenomics
Hate Polkadot parachain mechanism
Olympus Dao, Liquid Lido, flashbots – from these pieces they’ll construct Atom 2.0
Atom 2.0 will benefit from the growing value of the shared security – but there’s no requirement to use Atom token at all

Built a lot of public goods infrastructure

Will share Atom 2.0 at Cosmoverse in 12 days
Will make EIP-1559 look like a joke
Something like parachain auctions but aligning with new tokens appearing in ecosystem – something better than airdrop mechanism

If not working on Cosmos, what would they work on
Jack – individual chain scaling – thought what Solana was doing was cool, DB optimization / consensus optimization
Zaki – privacy, MEV, wants to see ETH rollup layers become peers to Cosmos Hub – feel like doing Cosmos again, wants to build awesome infrastructure

Jack – wants to see non-Cosmos chains using IBC

You don’t have to use Cosmos SDK to build Cosmos chains – examples include Penumbra, Anoma (sp?)

Every step of Tendermint consensus algorithm is programmable (ABC++?)

Unleash app chains to innovate, solve MEV

Jason – “All roads in blockchain lead back to MEV”

Zaki – wants to speak native IBC to Ethereum – instead of Gravity Bridge (related to Sommelier?)

Most misunderstood about Cosmos – like Ethereum, more of a developer collective, fractal relationships
Strange and wonderful characters
Every validator runs its own influencer biz
Funds ask him how to get onboard – it’s quite confusing – no agreed / simple way

Santi: Starbucks partnership with Polygon – Ryan Watt has been very good at bringing enterprises onboard; Like VHS vs Betamax debate – best tech doesn’t necessarily win
Zaki: would kill himself if he had to run the BD deal to bring Starbucks onboard – “existential crisis every day”
Polygon is fork of pre-Cosmos Hub launch of Tendermint

“If Atom 2.0 is not fire, you can bring me on and just yell at me” – Zaki

Podcast notes – Peter Thiel on Uncommon Knowledge – “When do we go from wisdom of crowds to madness of crowds?”

Peter Thiel – Stanford + Stanford Law, PayPal cofounder, first Facebook investor

No ticker tape parades in NYC for individuals in 21st century – now it’s for groups like “healthcare professionals”, before it was for individuals like Charles Lindbergh

Prevailing view is heterodox thought no longer allowed, scared of putting individuals on pedestal

Believes in classic libertarian values – but it can be somewhat cowardly way of saying you’re a loser, you wanna be left alone
Acceptable for him to support Ron Paul for president, much more dangerous to support Donald Trump

When do we go from wisdom of crowds to madness of crowds?

If you win 99% of election, you’re in North Korea – you haven’t arrived at absolute truth but you’re in an insane totalitarian place

West has 2 philosophical traditions – Greco Roman and Judeo Christian

Covid – all kinds of things were asserted too dogmatically – and took hairpin turns

Science fights 2 front war against excess skepticism and excess dogmatism
Science thinks of itself as more fighting against dogmatism – “choose your enemies well because you’ll soon be like them” – now “science” can seem like a dogma, post-modern

Case of Jay Bhattacharya – example of sheer insanity – tenured Stanford professor, libertarian and heterodox thinker
He wrote article saying “there’s no high quality studies that prove wearing masks is effective” – triggered people, crazed campus reaction
Nuanced nature made it dangerous – supposed to think in clean bright line ideological terms
If you’re not allowed to say something, he has suspicion that not only should you say it, it’s simply true

Fauci: “When people criticize me, they’re really criticizing science because I represent science”
Science has become quite opposite from the exploration, open-debate that it should be
Real science doesn’t need to be called “science” – real chemistry, physics, etc – but eg “climate science” is like a tell in poker, exaggerating because it’s not quite there
Science with capital S seems like antonym of science lower case S

20 years of telling ourselves lies about Afghanistan – that it’s going wonderful, we’re nation building, form of epistemic closure
Believes Trump had a fundamentally correct view of Afghanistan – Trump said “Afghanistan is a shit country” – not rigorous or nice thing to say, but “when you limit yourselves to saying something that’s very nice, you can’t actually talk about anything at all”

Political correctness as misdirected form of politeness

Saying anything you want to — but remain civil

Fed Reserve
Inflation is common sense and everyone can see it – gas bill, grocery bill
Another case of epistemic closure – Fed seemed like last institution to register that inflation was accelerating
One of our most sacred institutions

MMT – everyone’s gravitated towards it at precise moment it should be questioned and challenged
Economics risks being very politicized – can twist the answers and go in strange direction
Not very precise a science, but when you violate everything, things will eventually go wrong

The hour is late for fiat money – clear signal is Bitcoin / Satoshi Nakamoto
Bitcoin is revolutionary anti-fiat money thing, a late warning like Trump’s warning on Afghanistan

Broken market
Fed buying all the bonds
Inflation showing up in assets, crypto, art, stocks, bonds
If inflation is 6% and rates are 0%, that’s still 6% confiscation of your cash / earnings

Davos – he stopped attending – people are only there as representatives of corporations and governments – it hit him that there are no real individuals in the room
Davos as sense of global government, participate as part of larger structure
Wisdom of crowds – center left politically correct thing
Not a truth seeking place, not maximum surface area of debate
It’s a world with no dissenting views
Doesn’t believe we’re at end of history

China
Ray Dalio says it’s like family with strict parents
He frames China since 2013 as Putin is positive role model (vs when they used to believe Soviet Union was a negative role model)
China works for one individual – Mr. Xi
Doesn’t work for anyone else anymore
Jack Ma as remarkable example
Tech companies have been clobbered
No individuals allowed, no wealthy people allowed
Back to totalitarian playbook
Quite different from China in 70s, 80s
China like much worse version of Japan – for many decades it had great model of copying and catching up, and then hit the wall – China seems like stranger more dysfunctional version of Japan
For most part, China has not overtaken us
2021 – feels like China has gone haywire, more Marxist economy, totalitarian escalation
Friend said “Xi as best thing happening to West”

Why are we in such a collectivist moment (in America)?
Something went wrong, not an idiosyncratic thing, symptomatic of bigger problem
Should have ticker tape parade for Satoshi – so much healthier sign for country – just symbolic but symbols are important

Is renewable possible? Or is it just historic decline
He believes it’s not historicism, inevitable trends – it’s individuals that matter

Advice to new grads:
Can start companies
Shortage of talent in government
Society feels increasingly on auto-pilot — but he would take opposite bet

Podcast notes – Agustin Lebron (Jane Street trader) on Lunar Society

Guest: Agustin Lebron – author of book Laws of Trading

The more you trade the more you realize how hard it is to do well

Studied engineering in college
Former chip designer, online poker player
Went into finance – Jane Street

Recently started a crypto company – building protocol to improve crypto trading

Finance – jobs are more fungible, harder to explain why you work at a specific firm since most jobs are similar and primarily money-driven

If you want to be a trader, implicitly you’re saying you care most about money
But a lot of the job is having inherent curiosity about many things, and about enjoying game for its own sake

Jane Street would select against people with prior retail trading experience

Concept of domain is narrower than people understand it to be – Robinhood trading is even more different than you think of trading at a market maker, or a hedge fund

Takes 6-18 months to train a trader to be net-positive
Jane Street was mostly Socratic method – sit with senior trader and learn at desk
Now there’s more structured training, learning an iterative thinking process

Most common failure mode in tech hiring is hiring too much for skills, instead of for potential
One of best hiring arbitrages is to “go more junior” – hire younger and earlier in careers

Agustin wants to find smartest 0.1% of high school students around world – put them into bootcamps for 6 months, learn useful skills, provide high skill jobs to graduates
This could be a trillion dollar business

SBF @ FTX – Future Fund to scout for this kind of talent

Intelligence (“G” factor) strongly predicts outcomes across jobs, industries
Why brainteasers so common in tech hiring – proxy for IQ (even though explicit IQ testing for hiring is illegal in the US)
Great example – Wonderlic test for football

For job candidates:
One of most important things you can do is select your coworkers
During hiring, get good at evaluating your interviewers too

Sheepskin effect – last semester of college boosts earnings vastly more – because of signaling / certification

San Diego tech cos love to hire Intuit employees with 2-3 years experience, because Intuit does great training – but you don’t see the really good employees that Intuit retained

If Agustin was a regulator – would ban leveraged ETFs, further regulate all the volatility products

// Stopped ~45% of the way thru

Podcast notes – Ujwal Velagapudi (SMB buyer, vending machine biz) – Acquiring Minds podcast

Guest: Ujwal Velagapudi
Podcast: Acquiring Minds

Hometown – suburbs of Detroit (he’s 30-ish)

Bought and sold a dozen cars on CraigsList (CL) while in HS / college
Always made a profit, and learned negotiation and hustling through it

Commercial building for sale on CL – seemed a great deal, 75yo seller owner / occupier
Fire damaged red brick building – 7 Mile in Van Dyke
Was a year out of college
Did entire lease + purchase agreement (PA) himself
Grew this commercial portfolio to 12 units – mixed use, light industrial – at this point was earning ~$5K/mo

Still had day job – supply chain management in aerospace, quit in 2018
But always interested in passive income

Then explored SMBs – pizza, night club – all on CL – spent a lot of time on deals but they all fell through
Finally saw a bar for sale – near Red Wings stadium – rich history, but the arena was being demolished and relocated
Had been bartender in college, even though he didn’t drink
75 seat bar, no kitchen, did some food trucks + temp chefs – $100-200K gross, 1 full-time bartender
Put in LOI, 4 days later had a signed agreement + down payment
Purchasing by himself – knowing it was mostly for his education – “wasn’t for the money”
Hired completely new team eg bartenders – also on CL
Ultimately couldn’t extend the lease, owners wanted him to leave so they could demo the building
Was initially there every day after his 9-5 job – slowed over time
Owned it for 1.5 years, then liquidated everything he could

Also on CL – saw a gym for sale – Snap Fitness, hot gym in area
Personally loved sports, exercise
Bought with cash for $75K – closed deal in 3.5 weeks – was making $50K (so 1.5x multiple), but LA Fitness had just moved in next door
2900 sf, prime downtown location in small Detroit suburb, yoga studio, tanning bed, weight lifting
24/7, keycard
Sold for 2.5x purchase price to his gym manager – had started as a customer, became assistant GM and then GM for him, and eventually bought it over
Always prioritized happiness and efficiency over maximizing profits, outsourced a lot of work, used tech as much as possible (video cams, PoS, DocuSign, etc)
Only spent 10 hours total in that gym – managed it like it didn’t rely on him, relied heavily on financials / reports instead of micro managing

Bought an ecomm business – Amazon FBA selling hats, bags, gloves – negotiated the heck out of it, thought he got a good deal
Offered $180K, $90K down ($500K revenues?) – was a well-known online brokerage
Then he discovered it had a lien, because seller took loan from Amazon which was automatically deducted from earnings
Seller tried to hijack the account, got FBI involved
Seller was a felon with drug issues
Took a big hit, ultimately liquidated it on EmpireFlippers (good experience)

In 2018, multiple income streams, quit full-time job, spent a few months in Thailand, had gone for an entrepreneurship conference
Spent a few months on the beach, 4 hour work week, but “it was so boring”

Increasingly difficult to find good M&A deals on Craigslist – back then there was a lot of misclassified deals, 80 year old sellers
Now brokers are picking off everything, sellers more sophisticated, marketplaces more fragmented (eg, FB marketplace)

After ecomm, purchased minority ownership in consulting business for niche insurance software
Moved to Texas for it, 20 year old company, was acquired by Oracle
Had monopoly in a very niche business as systems integrator
But covid hit and really hurt business
Oracle refused to sell it back to him
Wasn’t able to compete with new startups

HiByron
MicroAcquire – saw a promising target, LOI in a few days, closed sale in 5-6 days
Acquired for 5-figures, $2.6K in MRR, US-based virtual assistants + tooling (geared toward online business
owners), $44K in trailing 12 mo revenue
Like a staffing agency + tech platform
Lot of R&D gone into it, great base, he was excited to own business
Great seller transition – SOP, tutorials – seller was focused on a new project

**Likes to close very quick, helps him w/ sellers“I’m a cash ready buyer that will close within 2 emails” – has standard set of questions, very attractive to sellers

Living in Mexico now

Let someone take over HiByron (the VA biz) – 60/40 split for any new revenues (“phantom equity”) + 1 year vesting
Has grown considerably with new management
From $2.6K MRR to $45K MRR – specifically one big new customer (a vendor to OnlyFans)

Latest acquisition: Vending business
Talked to a lot of brokers, shared requirements, looked in Mexico & Canada too
Looking for $multi-million deal
3 targets:
1. Chemical distribution (60 yo biz) – lost to strategic acquirer
2. Tech biz (phone systems securitization) – lost to PE / group of investors
3. Vending biz – jukebox, arcade games, ATMs – bars, restaurants – “seemed super weird to me”

Reviewed financials – very stable business – 7-figure cashflow, 50/50 split, labor is biggest OpEx
6 month process to close – all his previous M&A experience really helped, sought financing from 100+ banks and lenders, dealt with a lot of rejection
Main issue: Entirely local – had to be around his base, maybe at most a 2-3 hour radius
End consumer is B2C, but direct customers are B2B (eg, restaurant or theater)
80-90% shutdown during covid

Vending biz primer
Owns and places equipment
Sends own collector + technician for operations
Two critical elements – 1) service + 2) customers
Service – If jukebox goes down at 11pm, you must take care of it asap; if broken, order parts and replace in 1-2 days
Customers – gotta baby ‘em, one of few vendors that actually PAYS the customer – mutually beneficial, perfect alignment
Don’t wanna be 1+ hours away from customers
In vending, LOTS of one-man shows, small route that they do everything for, biz plateaus
A few are vending aggregators – buy multiple routes
Lots of vending sellers came to him after he bought the first one
Bought 3 more since that first purchase
His operation is biggest in the state, but doesn’t compare to national operators
Cash-flow is now multiple 7-figures

Podcast notes – Demis Hassabis (CEO of DeepMind) – Lex Fridman

This podcast made me feel very stupid, and very inspired.

Turing test – in 1950s, Turing didn’t mean it to be a rigorous formal test, more of a philosophy experiment, didn’t specify things like parameters of test, how long test should last, etc

More modalities than just language to express intelligence eg physical movement

Played chess at 4, earnings from winning a chess competition let him buy a computer
Bought programming books, started making games, felt they were a magical extension of your mind

“AI is ultimate expression of what a machine can do or learn”

At 12yo he got to chess masters level, the process of learning chess makes him think a lot about thinking and about brains

“Chess computer handbook” by David Levy – explained how chess programs were made

First AI program he built was on his Amiga, programmed Othello

Wrote game called “Theme Park” with a core AI, sandbox game, reacted to players, every game was unique

He designed and wrote AI for games in 90s – at the time, game industry was cutting edge of tech (GPUs for game graphics, AI, John Carmack)

“Black and white” game – train a pet, and depending on how you train it, it would be more or less kind to others, powerful example of reinforcement learning

DeepMind – core part of strategy from start was to use games to test how well AI is doing, if the ideas are working
Eg, Go – clear rules and win conditions, humans have played for thousands of years, easy to test how good is your system vs human players
Part of why their AI has progressed so quickly – by developing against games

“Chess is drosophila of intelligence” – Gary Kasparov
Many AI researchers have all wrote chess AI programs

DeepBlue beating Kasparov was a huge moment – he was in college at time – came away more impressed with Kasparov’s mind than with DeepBlue (because Kasparov could play almost at the AI level, but could also do all these other things as a human, while DeepBlue at that time couldn’t even play tic-tac-toe)

What makes chess compelling as a game?
Creative tension between bishop and knight – leads to a lot of dynamism
Chess has evolved to balance those two more or less equally (worth 3 points each)
Balanced by humanity over hundreds of years

Different levels of creativity
1. Lowest level is interpolation – averaging everything you see (eg, “an average looking cat”)
2. Next is extrapolation – AI coming up with a new move in Go that no one’s seen
3. Out of the box innovation – coming up with a new game entirely – AI nowhere close to this yet

Currently AI can do 1 and 2 but not 3
For 3, if you were to instruct an AI to create a game, you’d say “come up with a game that takes 5 minutes to learn, but lifetimes to master, aesthetically beautiful, and can be completed in 3-4 hours”
We can’t abstract high level notions like that to AIs (yet)

AI could be used to make current games better by taking game system, playing millions of times, and then improving the balance of rules and parameters – give it a base set + Monte Carlo tree search – takes humans many years and thousands of testers to do it

His first big game was theme park, amusement park – then whole cities – and Will Wright’s made SimEarth simulating the whole earth

“Simulation theory” – doesn’t believe it, in sense that we’re in a computer simulation / game, but does think best way to understand physics and universe is from computation perspective – information as fundamental unit of reality instead of energy or matter
Understanding physics as information theory could be valuable

Roger Penrose – Emperor’s New Mind – he believes we need quantum, something more, to explain consciousness in the brain
Most neuroscientists / mainstream biologists haven’t found any evidence of this
While continually classic Turing machines keep improving – and DeepMind / Demis work is champion of this
Thinks universal Turing machines can eventually mimic human brain without Penrose need for something more

Something profoundly beautiful and amazing about our brains – incredibly efficient machines, in awe of it
Building AI and comparing to human mind will help us unlock what’s truly unique about our minds – consciousness, dreaming, creativity
Philosophy of mind – there haven’t been the tools, but today we increasingly have them

Lex – Universe built human mind which built computers to help us understand universe and human minds

Protein folding – AlphaFold 2 solved it
Proteins = essential to all life – workhorses of biology, amazing bio-nano machines, specified by genetic sequence, in the body they fold into 3D structure (like string of beads folded into a ball)
The 3D structure determines what it does – and drugs must understand this to interact with it
Structure maps to function, and is specified by amino acid sequence
Unique mapping for every protein – but it’s not obvious – and almost infinite possibilities
Can you by studying the sequence, predict the 3D structure?
Takes 1 PhD student an entire PhD to predict one protein
But AlphaFold 2 can do it in seconds now – over Christmas can do it over entire human proteome space (!!)
Biologists can now lookup protein 3D structure in a google search

AlphaFold was most complex and meaningful system they’ve built so far
Started on games (AlphaGo, AlphaZero), to bootstrap general learning systems
His passion is scientific challenges – AlphaFold is first proof point
30 component algorithms needed to crack protein folding
About 150K protein structures had been found – that was their training set
Would put some of AF’s best predictions back into training set to accelerate training
AF2 was truly end to end – from amino acid sequence directly to 3D structure, without needing all the intermediary steps – system is better at learning the constraints on its own instead of guiding it

AlphaGo – learning system but trained only to learn Go
AlphaZero – removed need to learn from human games – just play with itself
MuZero – didn’t even need to give rules, just let it learn on its own

Started DeepMind in 2010 – back then no one was talking about AI, people mostly thought it doesn’t work (even at MIT)
If all professors tell you you’re mad, at least you know you’re on a unique track
Founding tenets / trends
-Algorithmic advances (reinforcement learning)
-Understanding about human brain (architectures, algos) improving
-Compute and GPUs improving
-Mathematical and theoretical definitions of intelligence

Early days – ideas were most important – deep reinforcement learning, transformers, scaling those up
As we get closer to AGI, engineering and data become more important
**For large models – scale is clearly necessary but perhaps not sufficient

DeepMind – purposely built multi-disciplinary organization – neuroscience + machine learning + mathematics + gaming, and now philosophers and ethicists too
“A new type of Bell Labs”
DeepMind itself is a learning machine building a learning machine

Top things to apply AI – biology and curing diseases (AlphaFold), but it’s just beginning
Eventually simulate a virtual cell (maybe in 10 years) – “that’s my dream”
Drugs take 10 years – target to drug candidate – maybe it can be shortened to 1 year with this, AlphaFold as first proof point

Math is perfect description language for physics
AI as perfect description language for biology (!)

Open-sourced AlphaFold (including data) – max benefit to humanity – so many downstream applications, better to accelerate research and discovery, used by 500K researchers (almost every biologist in the world!), amazing fundamental research, almost every pharma company is using it, “gateway drug to biology”

Also open-sourced MuJoCo – purchased it explicitly to open source it

One day an AI system could come up with something like General Relativity (!)

Big new breakthroughs will come at intersection of different subject areas (DeepMind = neuroscience + AI engineering)
We just don’t understand what it’d be like to hold the entire internet in your head (imagine reading all of Wikipedia, but much much greater) – no one knows what will result

Nuclear fusion – believe AI can help
In any new field, talk to domain experts for collaboration
What are all the bottleneck problems? Think from 1st principles
Which AI methods can help
Problem of plasma control is great example – plasma is unstable (mini-sun in a reactor), want to predict what plasma will do next, to best model and control it
They’ve largely solved it with AI, and now looking for other problems in fusion

Simulating properties of electrons – if you do it, you can describe how elements and materials work (fundamental to materials science)
Would like to simulate large materials – approximate Schrodinger’s equation

His ultimate aim for AI – to build a tool to help us understand the universe – to test the limits of physics
A true scientist – the more you find out, the more you realize you don’t know
Time, consciousness, gravity, life – fundamental things of nature – we don’t really know what they are
We treat them as fact and box them off – but there’s a lot of uncertainty about what it is
Use of AI is to accelerate science to the maximum – imagine a tree of all knowledge – we’ve barely scratched surface, and AI will turbocharge all of it – understanding and finding patterns, and then building tools

If you’re good at chess, you still can’t come up with a move like Garry Kasparov, but he can explain the move to you – potentially AI systems could understand things we could never by ourselves, and then explain it and make it useful for us

We’re already symbiotic with our phones and computers, Neuralink, and could augment / integrate with these AI

His current feeling is we are alone (no aliens)
We could easily be a million years ahead or behind in our evolution, eg, if meteor that destroyed dinosaurs came earlier or later – and in a few hundred years, imagine where we’ll be, AI, space traveling – we’ll be spreading across the stars; will only take ~1M years for Von Neumann systems to populate across the galaxy with that tech
We should have heard a cacophony of voices – but we haven’t heard anything
“We’ve searched enough – it should be everywhere”
If we’re alone, somewhat comforting re: Great Filter (maybe we’ve passed it)

Wouldn’t be surprised if we found single cell alien life – but multi-cellular seems incredibly hard
Another large leap is conscious intelligence
General intelligence is costly to begin with – 20% of body’s energy – a game of professional chess is same as F1 racer
Hard to justify evolutionarily – which is why it’s only been done once (on Earth)

AI systems – easy to craft specific solutions, but hard to do generally – at first general systems are way worse

Do AGI systems need consciousness?
Consciousness and intelligence are double dissociable – can have one without the other in both ways
Eg, Dogs have consciousness and self-aware but not very intelligent, most animals are pretty specialized
Eg, some AI are amazingly smart at playing certain games or doing certain tasks, but don’t seem conscious

May be our responsibility to build systems that are not sentient
None of our systems today have one iota of consciousness or sentience – way too premature
Re: Google engineer who believed their language system was sentient – Demis believes it’s more a projection of our own minds, our desire to construct narrative and agency even within inanimate systems
Eliza AI chat bots in 1960s – already fooled some people

Neuroscience – certain pre-reqs may be required for consciousness, like self-awareness, coherent preferences over time

Turing test is important, but there’s second that differ in machines: we’re not running on same substrate (humans are carbon based squishy life forms)

Language models – we don’t understand them well enough yet to deploy them at-scale
Should AI be required to announce that it is AI?

re: AI ethics, important to look at theology, philosophy, arts & humanities
Heading into an area of radical abundance and knowledge breakthroughs if we do it right – but also huge risks
Need careful controlled testing instead of just releasing into the wild, the harms could be fast and huge

Better to first build these AI systems as tools – carefully experiment and understand – before we really focus on sentience

How to prevent being corrupted by this AI power:
-Important to remain grounded and humble
-Being multi-disciplinary keeps you humble – because always better experts
-Have good colleagues who are also grounded

AI can learn for itself most of its knowledge, but will have residue of culture / values from who builds it

Globally we can’t seem to cooperate well eg, climate change
Need to remove scarcity to help promote world peace – radical abundance

AI should belong to the world and humanity, everyone should have a say

Advice for young
-what are your true passions? explore as many things as possible; find connections between things
-understand yourself – how do you deal with pressure, hone your uniqueness and skills

Perfect day in Demis’ life, habits
-10 years ago: whole day of research + programming, reading lots of papers, reading sci-fi at night or playing games
-today: very structured, complete night-owl, 11-7pm work (back to back meetings, meet as many people as possible), go home and spend time with family, 10pm-4am do individual work, long stretches of thinking and planning and some email
-quiet hours of morning – love that time (1-3am), inspiring music, think deep thoughts, read philosophy books, do all his creative thinking, get into flow, sometimes will go to 6am next day, and pay for it the next day (but it’s worth it)

Always been a generalist – too many interesting things to spend time on just one

Lex: Why are we here?
Demis: To gain knowledge and understand the universe, understand world around us, humanity, and all these things flow from it: compassion, self-knowledge

Feel like universe is almost structured to let itself be understood and learned – why are computers even possible?

If Demis could ask one question of true AGI: “What’s true nature of reality?”
Answer could be a more fundamental explanation of physics, and how to prove them out
A deeper, simpler explanation of things – leading to consciousness, dreaming, life, gravity