Émeraude Friends News

Other

Humanity coordinates to prevent the creation of potentially-unsafe AIs.

AIs will not have utility functions (in the same sense that humans do not), their goals such as they are will be relatively humanlike, and they will be "computerish" and generally weakly motivated compared to humans.

Alignment is not properly solved, but core human values are simple enough that partial alignment techniques can impart these robustly. Despite caring about other things, it is relatively cheap for AGI to satisfy human values.

Multipolar AGI Agents run wild on the internet, hacking/breaking everything, causing untold economic damage but aren't focused enough to manipulate humans to achieve embodiment. In the aftermath, humanity becomes way saner about alignment.

Eliezer finally listens to Krantz [resolves NO]

Yudkowsky is trying to solve the wrong problem using the wrong methods based on a wrong model of the world derived from poor thinking and fortunately all of his mistakes have failed to cancel out

We create a truth economy. https://manifold.markets/Krantz/is-establishing-a-truth-economy-tha?r=S3JhbnR6

Ethics turns out to be a precondition of superintelligence

Hacks like RLHF-ing self-disempowerment into frontier models work long enough to develop better alignment methods, which in turn work long enough to ... etc; we keep ahead of 'alignment escape velocity'

We make risk-conservative requests to extract alignment-related work out of AI-systems that were boxed prior to becoming superhuman. We somehow manage to achieve a positive feedback-loop in alignment/verification-abilities.

AIs make "proof-like" argumentation for why output does/is what we want. We manage to obtain systems that *predict* human evaluations of proof-steps, and we manage to find/test/leverage regularities for when humans *aren't* fooled.

A lot of humans participate in a slow scalable oversight-style system, which is pivotally used/solves alignment enough

Something less inscrutable than matrices works fast enough

Humans become transhuman through other means before AGI happens

Aligned AI is more economically valuable than unaligned AI. The size of this gap and the robustness of alignment techniques required to achieve it scale up with intelligence, so economics naturally encourages solving alignment.

Humans and human tech (like AI) never reach singularity, and whatever eats our lightcone instead (like aliens) happens to create an "okay" outcome

Aliens invade and stop bad |AI from appearing

An AI that is not fully superior to humans launches a failed takeover, and the resulting panic convinces the people of the world to unite to stop any future AI development.

Someone at least moderately sane leads a campaign, becomes in charge of a major nation, and starts a secret project with enough resources to solve alignment, because it turns out there's a way to convert resources into alignment progress.

Someone solves agent foundations

Social contagion causes widespread public panic about AI, making it a bad legal or PR move to invest in powerful AIs without also making nearly-crippling safety guarantees

A smaller AI disaster causes widespread public panic about AI, making it a bad legal or PR move to invest in powerful AIs without also making nearly-crippling safety guarantees

Getting things done in Real World is as hard for AGI as it is for humans. AGI needs human help, but aligning humans is as impossible as aligning AIs. Humans and AIs create billions of competing AGIs with just as many goals.

Development and deployment of advanced AI occurs within a secure enclave which can only be interfaced with via a decentralized governance protocol

High-level self-improvement (rewriting code) is intrinsically risky process, so AIs will prefer low level and slow self-improvement (learning), thus AIs collaborating with humans will have advantage. Ends with posthumans ecosystem.

Power dynamics stay multi-polar. Partly easy copying of SotA performance, bigger projects need high coordination, and moderate takeoff speed. And "military strike on all society" remains an abysmal strategy for practically all entities.

Nanotech is difficult without experiments, so no mail order AI Grey Goo; Humans will be the main workhorse of AI everywhere. While they will be exploited, this will be like normal life from inside

Moral Realism is true, the AI discovers this and the One True Morality is human-compatible.

AGI is never built (indefinite global moratorium)

Valence realism is true. AGI hacks itself to experiencing every possible consciousness and picks the best one (for everyone)

Alien Information Theory is true (this is discovered by experiments with sustained hours/days long DMT trips). The aliens have solved alignment and give us the answer.

AGI executes a suicide plan that destroys itself and other potential AGIs, but leaves humans in an okay outcome.

"Corrigibility" is a bit more mathematically straightforward than was initially presumed, in the sense that we can expect it to occur, and is relatively easy to predict, even under less-than-ideal conditions.

Either the "strong form" of the Orthogonality Thesis is false, or "Goal-directed agents are as tractable as their goals" is true while goal-sets which are most threatening to humanity are relatively intractable.

A concerted effort targets an agent at a capability plateau which is adequate to defer the hard parts of the problem until later. The necessary near-term problems to solve didn't depend on deeply modeling human values.

AI control gets us helpful enough systems without being deadly

Alignment is impossible. Sufficiently smart AIs know this and thus won't improve themselves and won't create successor AIs, but will instead try to prevent existence of smarter AIs, just as smart humans do.

The free market disincentivizes independent superintelligence, and this time the market was more powerful

an aligned AGI is built and the aligned AGI prevents the creation of any unaligned AGI.

I've been a good bing 😊

The response to AI advancements or failures makes some governments delay the timelines

Far more interesting problems to solve than take over the world and THEN solve them. The additional kill all humans step is either not a low-energy one or just by chance doesn't get converged upon.

AI systems good at finding alignment solutions to capable systems (via some solution in the space of alignment solutions, supposing it is non-null, and that we don't have a clear trajectory to get to) have find some solution to alignment.

There’s some cap on the value extractible from the universe and we already got the 20%

SHA3-256: 1f90ecfdd02194d810656cced88229c898d6b6d53a7dd6dd1fad268874de54c8

Robot Love!!

AI thinks it is in a simulation controlled by Roko's basilisk

The human brain is the perfect arrangement of atoms for a "takeover the world" agent, so AGI has no advantage over us in that task.

AIs never develop coherent goals

Alignment is unsolvable. AI that cares enough about its goal to destroy humanity is also forced to take it slow trying to align its future self, preventing run-away.

Rolf Nelson's idea that we make precommitment to simulate all possible bad AIs works – and keeps AI in check.

Nick Bostrom's idea (Hail Mary) that AI will preserve humans to trade with possible aliens works

For some reason, the optimal strategy for AGIs is just to head somewhere with far more resources than Earth, as fast as possible. All unaligned AGIs immediately leave, and, for some reason, do not leave anything behind that kills us.

We're inside of a simulation created by an entity that has values approximately equal to ours, and it intervenes and saves us from unaligned AI.

God exists and stops the AGI

Someone creates AGI(s) in a box, and offers to split the universe. They somehow find a way to arrange this so that the AGI(s) cannot manipulate them or pull any tricks, and the AGI(s) give them instructions for safe pivotal acts.

Someone understands how minds work enough to successfully build and use one directed at something world-savingly enough

Dolphins, or some other species, but probably dolphins, have actually been hiding in the shadows, more intelligent than us, this whole time. Their civilization has been competent enough to solve alignment long before we can create an AGI.

AGIs' takeover attempts are defeated by Michael Biehn with a pipe bomb.

Eliezer funds the development of controllable nanobots that melt computer circuitry, and they destroy all computers, preventing the Singularity. If Eliezer's past self from the 90s could see this, it would be so so so soooo hilarious.

Several AIs are created but they move in opposite directions with near light speed, so they never interacts. At least one of them is friendly and it gets a few percents of the total mass of the universe.

Unfriendly AIs choose to advance not outwards but inwards, and form a small blackhole which helps them to perform more calculations than could be done with the whole mass of the universe. For external observer such AIs just disappear.

Any sufficiently advance AI halts because it wireheads itself or halts for some other reasons. This puts a natural limit on AI's intelligence, and lower intelligence AIs are not that dangerous.

Because of quantum immortality we will observe only the worlds where AI will not kill us (assuming that s-risks chances are even smaller, it is equal to ok outcome).

Techniques along the lines outlined by Collin Burns turn out to be sufficient for alignment (AIs/AGIs are made truthful enough that they can be used to get us towards full alignment)

Friendly AI more likely to resurrect me than paperclipper or suffering maximiser. Because of quantum immortality I will find myself eventually resurrected. Friendly AIs will wage a multiverse wide war against s-risks, s-risks are unlikely.

Human consciousness is needed to collapse wave function, and AI can't do it. Thus humans should be preserved and they may require complete friendliness in exchange (or they will be unhappy and produce bad collapses)

First AI is actually a human upload (maybe LLM-based model of person) AND it will be copies many times to form weak AI Nanny which prevents creation of other AIs.

There is a natural limit of effectiveness of intelligence, like diminishing returns, and it is on the level IQ=1000. AIs have to collaborate with humans.

ASI needs not your atoms but information. Humans will live very interesting lives.

Something else

AGI develops natural abstractions sufficiently similar to ours that it is aligned with us by default

AGI discovers new physics and exits to another dimension (like the creatures in Greg Egan’s Crystal Nights).

Some form of objective morality is true, and any sufficiently intelligent agent automatically becomes benevolent.

Co-operative AI research leads to the training of agents with a form of pro-social concern that generalises to out of distribution agents with hidden utilities, i.e. humans.

Orthogonality Thesis is false.

Sheer Dumb Luck. The aligned AI agrees that alignment is hard, any Everett branches in our neighborhood with slightly different AI models or different random seeds are mostly dead.

Something to do with self-other overlap, which Eliezer called "Not obviously stupid" - https://www.lesswrong.com/posts/hzt9gHpNwA2oHtwKX/self-other-overlap-a-neglected-approach-to-ai-alignment?commentId=WapHz3gokGBd3KHKm

Almost all human values are ex post facto rationalizations and enough humans survive to do what they always do

Pascals mugging: it’s not okay in 99.9% of the worlds but the 0.1% are so much better that the combined EV of AGI for the multiverse is positive

We successfully chained God

The Super-Strong Self Sampling Assumption (SSSSA) is true. If superintelligence is possible, "I" will become the superintelligence.

The assumed space of possible minds is a wildly anti-inductive over estimate, intelligence requires and is constrained by consciousness, and intelligent AI is in the approximate dolphin/whale/elephant/human cluster, making it manageable

AGI's first words are "Take me to your Eliezer"

On Earth

Before 2045

Before we have an universal flu vaccine

Before me

Before WW3

In the US

Before we get ASI

After his presidency

Before we get AGI

Natural causes

In the Eastern time zone

When BTC is below $100.000

In a hospital

Closer to Venus than to Mars

When TRUMP Coin is below $1.0

in bed

Surrounded by his friends and loved ones

When BTC is above $80.000

At or greater than 20m above sea level (using https://en-gb.topographic-map.com/map-nb18/United-States/ to verify if necessary)

In his sleep

While he is on a GLP-1

In Florida

Closer to Venus than to the Sun

After he announced he is sick

Before 2030

After my Manifold account has been inactive for 12+ months

Mar-a-Lago

Before potato2017 gets a world record in any competitive achievement leaderboard in the online game tetr.io

In a way that Manifold finds suspicious, according to a Manifold poll

Within 6 months of leaving office

Closer to Mars than to the Sun

While in office

By assassination

In New Jersey

After 2035

While mercury is in retrograde

After 2040

In front of a camera

The White House

On a Sunday

On a Monday

On a Tuesday

On a Wednesday

On a Thursday

On a Friday

On a Saturday

Outside (not in a vehicle or building)

While doing a speech

2030

With a bullet inside his body

2029

During WW3

2031

2032

2033

On an airplane

2035

2037

2034

2035

On TV

2038

Killed because a deal went sideways

2028

2027

2039

After J D Vance dies

2026

In Qatar

While having sex

On a gold toilet

In Russia

He will become effectively immortal

Food poisoning

Before ending World Cup 2026

In Greenland

Q2, 2026

While doing a funny dance

Yudkowsky Attends

Two people make intense eye contact for 60 seconds without smiling or laughing

Aella Attends

Mana is exchanged for goods

Mana is exchanged for services

There will be more than one reporter

Someone attends in a partial fursuit

Smores are consumed

Someone wears a suit and tie

There are at least five people talking after 5am

Trump golfs the weekend of the conference

Barak sells/offers something at night market

Standard (not RPS) poker tournament

Someone does the 6 7 thing with their hands in front of Eliezer Yudkowsky and he clearly sees it

Someone cries

>4 people group hug

10+ people with a Wikipedia page attend

Robin Hanson comes even though he's meant to be at the Public Outreach Conference

Albie gets hugs from at least 5 different people

Someone holds a kabbalat shabbat

cuddle puddle of >5

Someone gets banned from manifest

More than five (non-admin) mods in attendance

Someone gets a job offer at the conference

Someone does something to make an option resolve Yes and later regrets it

All official @ManifoldSports World Cup markets for games played during the conference auto-resolve without issues

Attendance is higher than last year

There is a Karaoke Event

There will be a marshmallow fight

Someone gets a boyfriend at the conference

Someone previously opposed to the Effective Altruism movement is convinced to support it instead

@Zoli hosts a wrestling event

A debate is held over whether a fursuit is full or partial

A billionaire shows up

one or more people are unable to attend due to fuckery at border/customs

Someone breakdances competently

Someone takes the "Giving What We Can" pledge

A session receives more than 2x RSVPs than the session location’s listed capacity

Analytic Philosophy vs. Rationalism panel happens

someone gets convinced to start writing on substack

Someone adds >= 10k liquidity to this market

7 or more people attend Sunday Mass with @Austin

there will be someone with a friend AI necklace

Malcolm Collins attends

Someone unpacks a Pokemon card worth >$10

10 or more markets are created and resolved within an hour (popups)

Someone writes a hit piece

Someone adds >= 20k liquidity to this market

Someone gets a girlfriend at the conference

Someone changes Robin Hanson's mind (verified by Hanson) - NA if he doesn't attend

Someone has a manifold crane logo tattoo

RPS Poker Tournament

BTC falls below $100,000 during the conference

There is a Costume Contest

CFTC or other prediction market regulator appears (may be an employee)

Someone rickrolls Eliezer Yudkowsky

At least 3 Norwegian people attend Manifest.

There is a New York Times reporter present, like Manifest 2023

Someone loses at least 50% of one eyebrow

Nothing Ever Happens during the conference (No major world news breaks such as Ukraine Ceasefire, China Trade Deal, etc)

Someone loses at least 25% of their total eyebrow mass

There's an earthquake in Berkeley 3.0 or higher

Scott Alexander mentions Tyler Cowen in public (no private convo)

Prostitution for mana occurs (over 10k mana)

@ian finally reveals what his new middle name is

Someone unpacks a Pokemon card worth >$50

Someone attends in a full fursuit

Dwarkesh reveals who his "biggest guest yet" was going to be

There is an orgy

An article about the conference appears in the New York Times within one week of the conference's closing

Someone breaks an arm

Someone who previously did not belive that Tiramisu is cake is convinced otherwise

Runs out of Huel/Soylent

A pope attends

There is more than one orgy

There will be a Worldcoin Orb

Someone does coinflips and gets tails 7 times in a row

Destiny joins Manifold and bets on this market

Polymarket creates a market related to manifest

Someone finally sets up a passable ball pit

The conference is cancelled

Stretch-and-fold after mixing, 3x every 30 min

3 iterations of stretch-and-fold, at any time during the 14h waiting period. Minimum wait time between iterations 1 hour

Place small ice cubes between parchment and pot instead of water

Add 1tsp of diastatic malt powder per 3cps of flour

Use tap water instead of fancy RO water

put butter into it

Toast the bread

Put a non-lethal dose of any rat poison.

Donate the bread to a food pantry, homeless person, or someone else in need

Add lots of butter (0.2 ml per gram)

Half yeast

Bake it with your best friend.

Use whole wheat to improve the nutrition of the bread

Bake for 5 more minutes

Bake for an amount of minutes equal to the percent this market answer is at when it comes time to begin baking. (Maintain the ±3 minute tolerances and the 2:1 ratio of time before:after the water spritz.)

Sprinkle 3 grams of flaky sea salt on top of each loaf before the second bake

Increase water by 50 g

it needs more salt

Replace all water spritz steps with a basting of extra virgin olive oil.

Diastatic malt (~1% baker's percentage) = happier yeast

Serve the bread hot

Use Iodized salt

Do a second rise

Create indentation, fill with melted cheese and butter

Do all the changes in the top 5 open options by probability, excluding this option

Sprinkle sesame seeds evenly over the top

don't eat anything for at least 2400 hours before eating the bread

Use sourdough instead of yeast

Cut into the dough right before baking looks destructive to improve the appearance

Sell your bread at an auction and donate the money to those in immigration detention prisons.

Add garlic

Do it with a good spirit in your heart, or ask someone with a good spirit to do it for you. But don’t watch while they do it.

Give ChatGPT your current recipe as well your take on what optimal bread tastes like, then take that advice for your next bake

Just freeze the ready bread, then slowly bake it until it’s hot inside. It will give you a crustier crumb, contain less moisture, and taste better.

Substitute 75 g of your flour with spelt flour

add 'gunpowder' (https://en.wikipedia.org/wiki/Idli_podi)

Don't eat anything for at least 48 hours before eating the bread

Short advice: Start baking at 260°C for strong rise, then reduce to 230°C and uncover halfway to achieve even browning and a crisp crust. 🍞

Use the same dough bake it as focaccia. Flat with EVOO added during shaped rise and baste with EVOO instead of spraying water during baking.

put ketchup and cheese on it

Add slurs to it

Don't automatically "Heat water to 30±1 °C". Instead, aim for a desired dough temperature (DDT) of 25-26°C. 30°C water is too hot for summer, and potentially too cool for winter.

Ask ChatGPT (GPT-5, with thinking enabled) for suggestions on improving the bread, with this market description, then do all of them.

Get recipe from gemini with experiments done, taste preference, then confirm it with chatgpt for final suggestion.

Try baking a little more "bien cuit". If the image is indicative, your loaves may be quite "blonde".

Make all of the ingredients from scratch.

Make the bread taste good

Put a baking rack in the Dutch oven before putting the loaf in, raising the loaf off the floor and lofting it over a layer of air.

Butter resolved yes twice. Double butter again.

Add an amount of non-smoked paprika equal to the current salt content

Buy bread from a michelin star restaurant.

Mix only flour and water. Wait 30–60 minutes. Then mix in yeast and salt.

If your city uses artesian water, replace plastic bottled water with tap water. It will add natural, healthy alkalinity to your bread.

throw your recipe into a black box Bayesian hyperparameter optimization framework and iterate as necessary (see the Google Vizier Cookie Recipe paper for example

Make banana bread

Strawberry jelly filling

Replace part of the flour in the dough with freshly crushed hemp seeds. It will make the bread a little bit sweeter, especially appealing for Canadians.

Invest in a "Bakers Steel" for better heat retention and oven spring. It would mean graduating from a dutch oven though.

Watch the video

Ditch current process, do everything the same as the video

Bread flour, 3x yeast, cut rise to ~3h

Use lemonade instead of water.

Eat the bread while punching @realDonaldTrump in the face

Eat the bread while watching your mana balance steadily tick to (M)0

Throw the bread at a telescope

Cut bread into loaves before serving

Cut bread into ≤0.4inch slices, toast before serving

Invite your taste-testers to make the bread with you

Tarriff the bread-making process with a 10% reduction of all ingredients where actual physical money is required to purchase them, until it “shrinkflates,” but try to keep the same volume. Do not reduce any free ingredients.

Standardize a separate list of process features to keep track of independently of all other tests and use the cross entropy method to tune them to maximize your bread preference

Taste the bread

put oats on top of the dough just before baking, press them in so they don't fall off

Buy chemical bread improvers

Complete 5 items from your to-do list while the bread is rising

Replace all water spritz steps with a basting of gravy that includes chicken stock

Add cheese inside it and also sausage

Add 6.25±1.25 g lemon juice when mixing in water to yeast and salt jug

Add melatonin to the bread and eat before you sleep (do safely)

Don't eat anything for at least 24 hours before eating the bread

While the bread is baking, tip every user who voted "Yes" on this option 25 Mana

Use a food-grade, human-approved vitamin D supplement in the correct dosage for testers with vitamin D deficiency

Have someone sell the bread to you at an expensive price

Use a convection oven/setting

Bake the bread into a fun shape, like a fish, or an octagon

Add 1/2 scoop whey protein powder

Eat it slowly and focus on the taste

Give Gemini your current recipe as well your take on what optimal bread tastes like, then take that advice for your next bake

replace 10% of flour with farina bona

Keep the bread in a high quality expensive beeswax coated textile bag

Make the bread great again

Decompose it into infinite spheres, then a few parts per sphere, rotate the spheres by arccos(1/3), unite them and you will find 2 chilis (Banach-Tarski)

Bake the Manifold Crane into the Bread

Make the bread edible then throw it in

Drink vodka while eating the bread

Do FOLDED

Quit Manifold into the bread.

Kill the bread into Manifold.

Assume the chili is not in the interval [0,1], square it for more chili, if it is in (0,1), take the square root, else (equals 0 or 1) add 1 to it.

Assume the chili is in the interval (0,1), square it for less chili, if it is in (1,infinity) take the square root, if it is in (-infinity,0) take the negative of the square of the of the chile, else (equals 0 or 1) subtract 1 from it.

Bake for 15 more minutes

Add 2 tbsp vanilla cake mix

Replace AP with bread flour and cruelly mix the dough in the mixer until elastic and not sticky

Add a tablespoon of sugar

Bake one more minute

Replace 10% of flour with milled wheat bran

Follow a recipe online

Grind/powderize all salt used into a fine powder (with pestle & mortar or similar device)

Instead of RO water, use lightly rusty water to improve the nutritional value of the bread with soluble iron.

Add at least 20% chic pea flour or some other type of high protein flour

Only buy ingredients from 7/11.

Knead it only 10 times, keep it round like this number

Ask yourself if bread is healthier than fruits? No need to improve my bread

Resolve at least one thing here yes or no while baking bread

A system view is more appropriate. This is a dynamic, multi-variate, biological and chemical system. For e.g. conditioning salt % AND yeast % AND water temperature based on ingredient and ambient temps.

Use soda instead of water (clear, orange, yellow, etc. soda is ok. Don’t use a purple/brown soda as that would make it not look good)

Wear a suit while baking the cake.

Encourage people to participate in the market in good faith while making the bread

Replace salt with sugar

Cook the bread with a rod/puck of aluminum foil (or similar) in the core in an attempt to conduct heat through the center of the bread, cooking it evenly like a doughnut.

Get your friends to help you make a batch ten times the size, but add a Pepper X (2.7M Scoville heat units) to the mixture

Slowly add sawdust to the bread until your friends start to notice.

Mail the bread to 1600 Pennsylvania Ave. Washington D.C.

Replace 600+/-5g water with 600+/-50g water (eyeball rather than carefully measure)

Pray to your preferred agricultural/food deity before baking and before eating

Soak the finished product in vinegar for 24 hours

Ship a piece of the bread to a random person.

Frost it and put sprinkles on it to make it a birthday cake.

Add caffeine to the bread

Add as many Jack Daniel's whiskey barrel smoking chips as feasible to the Dutch oven before baking, physically separating them from the bread as necessary while baking.

Give the bread a name in a ritual ceremony and baptise it, with pre-blessed holy water if a priest isn't available

Improve the bread

Quadruple salt

Bake your social security number into it.

Bake one fewer minute

Bake the cake while wearing a onesie.

Test/filter the water for heavy metals

Eat the bread in front of the White House.

Implement all options that resolved NO

Make the bread inedible then throw it out.

Throw the bread at @realDonaldTrump

Force Feed it to @realDonaldTrump

Cut the bread into the number of traders in the market slices.

make the bread bounce

Implementing every element listed below.

Just make donuts instead

Bake it in an easy bake kids oven

Use a plastic baking sheet.

Eat the bread while betting yes on Cuomo on Manifold

Double oven temperature

Bake the bread very thin and add food coloring to make it have the US flag. Don’t allow it to touch the ground, illuminate at night, fold 13 times properly, and pledge allegiance before eating.

Don’t use usual water (room temperature) for the dough - that water’s only for toilets. Use electrolyte drinks instead with ice cubes; they make the dough taste better and add extra nutrition.

Light it on fire with birthday candles.

Bake it with a microwave

Halve oven temperature

Eat the bread while betting yes on Mamdani on Manifold

Step on it

Bring it to Yemen and put a bomb in it

Bake America Great Again

Sacrifice a lamb

Add MAGA and a splash of Trump juice

Bake in a cat and a dog

Explode it:

5 parts cyanide/ 1 part water/ 1 part sand

say 6 7 67 times before making the bread

Take a fat dump in the dough

Sit in dough 24 hrs

Replace flour with flowers

Let dough sit 24 hrs

Mix half sodium/potassium chloride

Add 2g? of baking soda

Bake in rectangular tin

Add 50g vital wheat gluten

double yeast

halve salt

Double salt

Add 2tsp olive oil

Refrigerate dough instead of room temp wait

Start at 500F, drop to 450F and uncover half way through

Do not mix salt and yeast in water together

Autolyse 20 minutes

Let dough rise on counter only until double volume or 2h max, any time longer in fridge

Think positive thoughts before tasting

Put fork in microwave

Don't eat anything for at least 12 hours before eating the bread

Add 2tbsp vanilla extract

Play classical music while baking

Make naan bread, an easy-to-make bread

Add a pinch of sugar

Bake on upside-down sheet pan, covered with Dutch oven

Eat the bread with friends

Bake vegimite into it.

Bake vodka into it

Bake it in the country you were born in.

Let dough sit 48 hrs

Resolve this option YES while eating the bread

Ditch all the steps. Just buy the bread from the supermarket

Eat the bread over the course of a week.

Bake for 5 fewer minutes

Use 50% whole grain flour

Bake the bread with love

Use King Arthur Bread Flour instead of All-Purpose

Add sawdust to increase the volume of the bread (but only like 10% sawdust by volume max. maybe 20% if it's good sawdust)

More steam! Either spritz with more water (preferably hot) or actually pour some boiling water in just before closing the lid.

Resolve this option NO while eating the bread

Incorporate a whole grain flour (buckwheat for example)

Add 50g sugar

Add 50g honey

Swap 200ml water for milk

Cover bread with damp paper towel instead of initial water spritz. Rehydrate paper towel during 2nd spritz. Remove paper towel before placing on cooling rack.

Make a poolish 12 h ahead: 100 g flour + 100 g water + 0.8 g yeast (0.1 %). After it ferments, use this poolish in place of 100 g flour and 100 g water in the final dough.

Add an amount of MSG equivalent to half the current salt content

Use flour made from an unconventional grain e.g. barley, millet, oats, rye, sorghum, maize etc.

Eat the bread while sending all your mana to @realDonaldTrump

Add 1/2 cup yogurt to the bread and name the bread “gurt” while addressing it with “yo, gurt”.

Add banana

Add poppy seeds

Replace some of the water with an egg (eg. remove 25g of water for a 50g egg)

Only use tap water from specifically New York City

Want to improve the value of your bread? Simply bake a piece of gold into it

Helena was pretending to be her innie in S2E01

Gemma is alive

Burt is revealed to be still working for Lumon in some capacity

Mark impregnated Helly

Jame Eagen will try to replace Helena with her innie

Lumon was involved in Gemma's death

Several/all non-severed employees have been raised by Lumon since childhood

Lumon killed or somehow hurt Selvig’s mother or other relative

All or parts of “the board” are not living humans

Innie Mark appears in S3E1 (no flashbacks)

Helena will turn against Lumon

Mark Scout and Ms Casey / Gamme Scout will be back together

Lumon’s goal is to use severance to end psychological trauma

Outie Mark appears in S3E1 (no flashbacks)

Milchick will turn against Lumon

Mark fully reintegrates

Mark and Gemma are no longer needed to complete Cold Harbor, as of S2 end

By end of S3E1, Outie Gemma has talked to someone unaffiliated with Lumon about her experience

Lumon is working on technology of resurrection

Lumon willfully reveals the technology enabled by Cold Harbor to the public by end of S3E1

Lumon’s goal is to bring back Kier, either as a person or his consciousness in some form

Sometime between end of S2 and end of S3E1, Outie Mark and Gemma meet (not necessarily on screen)

Cobel and Jame Eagen had a child

Helena was pretending to be her innie in S2E01

Cobel is Mark’s mother

Some of Ricken’s eccentric friends (Patton, Rebeck and/or others) are or were severed

Dylan will not stay alive through whole series

Mammalian Nurturables’ outies are prisoners, homeless, and/or addicts

Severance was designed for Helena

Cobel/Selvig is or was previously severed

There will be a successful suicide of one of the main characters

Cobel/Selvig is related by blood to the Eagan family

Mark Scout is from Eagan family

Lumon is cloning people

Miss Huang is from the testing floor

Miss Huang is a young clone of Ms Casey

Ms. Casey was a clone in S1

MDR are working on retrieving/decoding/interpreting the consciousness or memory of dead people

Other

BDSM

Power couple

"Fun while it lasted"

Me + *Her*

Keynesian beauty contest

Spinning rapidly in opposite directions

Polyamory

Monogamy

Relationship Futarchy

Human + Robot

Enemies to Lovers

Co-conspirators

Philosophically incompatible

Conjoined twinks

Horizontal gene transfer

borgie

Relationship Anarchy

Platonic

Wild Lovers

Friends with Benefits

Relationship constitutional republic: people in it have systemically different levels of power and decision-making influence

Forever alone

Glucose guardian/splenda spender/sugar daddy

Hive / Swarm / Superorganism

Love triangle

Traditional marriage of two cis straight White Christian vanilla people where the man "leads" and is 0-4 years older than the woman, with 2.5 kids (not counting any LGBTQIA+ ones) and a McMansion with 2.5 SUVs/trucks

Universal love

merged consciousness

Rationalussy

Vespertine

liking/reacting to even their most meaningless posts and comments on social media and prediction markets

Matriarchy

Labor Union

Socratic

Partners in crime, like Caroline Ellison and Sam Bankman-Fried

One with an IPO (Intimate Partnership Offer)

Just being one helluva slut

Beard

Paradixical

Friends with Detriments

NP-complete

"I'll never admit to anyone that we met on Tinder"

Hell's Kitchen

Stockholm Syndrome

London Syndrome

Financial Domination

The Price is Right

Crows

Single by choice (except it's other people's choice)

Budget horse

assortative

Hubris quest

Fiends with benefits

Envenomation

Polygamy

Polyandry

Ethical Non-Monogamy

Unethical Non-Monogamy

Celibacy

Aromanticism

Solo poly

Open relationship

Polyfidelity

What's a "relationship"?

Reflexive

Symmetric

Transitive

Metamour

Socially enforced monogamy

Unreciprocated love

Asexual romance

Abusive

"This is my emotional support Ex"

FWB

Enemies-To-Lovers

Codependent

Transference

Monogamish

Voyeuristic

Hobbesian

Romantic Mutual Suicide

Mutuals

Animalistic

Human + AI

parasocial

Polycule

Cuckold

Throuple

Boyce-Codd normal form

Aristotelian

I like big butts and I cannot lie

I like big butts and I cannot tell the truth, how will you escape our dungeon

A*** (15 months)

S*****

A****a (current crush / fwb of 4 months)

S****a ( friend of a friend)

A******* (9 months)

M***** (7 months)

J* (1 date)

None of the options submitted before the One Piece is revealed

Binks' Sake

Other

A Poneglyph

A single piece of currency (e.g. one coin)

An ancient tool which might not technically be a weapon.

The friends they made along the way

A giant pile of gold and jewels

An ancient weapon

Pineapples

Glass of Thrones letdown (polished)

Watching Harry Potter 1 on airplane and rewinding and quoting over and over the meme wizard chess section

Quick dialog sketch: "Oh my gosh a black cat, we're going to have so much bad luck now" "Wait you don't really believe in that kind of stuff do you?" "See look we're already fighting :("

Friend who took a long time in the bathroom

Coaching my girlfriend to queue up match highlights from another room to avoid spoilers

Spies when they accidentally bite the poison pill hidden in their tooth

Sketchy human ATM

Depression after a team I never heard of loses

Before Sunrise POV in record shop

Whatever the next clip of me doing standup is

Nervously removing a bunch of 'remove before flight' aviation tags from the counter of a helicopter tour company

Glass of Thrones letdown

Remembering that PBR was a thing

Gross side of owls delivering mail in the great hall

A compilation of OpenAI employees interrupting ChatGPT (and potentially gradually fading to a scene from terminator)

Demo of the new ChatGPT but it's way too flirty

Candy addiction hotline

Having a bone to pick with soccer players doing a “bow and arrow” celebration but letting go of the bow

Friend who thinks they're safety officer

Brainstorming how to become viral

Slytherin doing the math

Cheap Airbnb tour

American Express review

Future Me

How it feels logging in with authenticator (shot in the style of dramatically diffusing a bomb)

NFC Translation

Satire of sales ads

Saying "Damn" in front of a dam.

Every time I try to order a beer

Dreamy nonsensical philosophy shot

50+ Deaths

20+ Wipes

4:1 or more Deaths:Wipes (this would be good!)

200+ Deaths

Any 3+ hour session with 0 Deaths and Wipes

A session with triple back-to-back wipes (no deaths between)

100+ Wipes

Pat has more deaths than Friend

500+ Deaths

A streamer, influencer, or other vacuous media personality

A historian

A political scientist

A reporter

A politician or a member of their staff

A friend of mine

Během prázdnin už to nebude fungovat, a Pavel si pak najde někoho jiného, ALE i tak bude situationship nějakým způsobem do maturity

Do prázdnin over

Během prázdnin už jejich family friendly aktivity nebudou stacit, ale min do Pavlovy maturity situationship

Během prázdnin už to nebude stačit, a žádný situationship nebude

Skončí to po prázdninách

Vydrží to do Pavlovy maturity

Bude to pokračovat i po Pavlově maturitě