
Sign up to save your podcasts
Or


At a petrol station you connect the hose and watch the numbers climb. So why can't Starship do the same thing in orbit? The pipe isn't really the puzzle. The puzzle is everything an ordinary petrol station quietly does for us.
This episode takes a car into orbit and removes those conveniences one at a time: matching motion to meet safely, finding the liquid when a tank has no ordinary bottom, using pressure to drive the flow, keeping methane and oxygen cold when space isn't a fridge, chilling down warm plumbing, working out who fills the filling station, and what tests so far have and haven't shown.
0:00 Why is refuelling in space hard?
0:42 We already refuel in space
1:27 Two ingredients
2:19 Match the motion
3:15 Connect, seal, check
4:02 Where is the bottom?
4:57 Settle toward the outlet
5:48 Pressure drives the flow
6:43 The gas matters too
7:33 Very cold liquids
8:13 Space is not a fridge
9:15 Vapour is not lost mass
10:07 First, cool the plumbing
10:58 Filling changes the machine
11:52 Who fills the station?
12:57 Divide the journey
13:57 Safe enough to repeat
14:47 Different tests, different proof
15:52 The hose is the visible bit
Instagram, X and TikTok: @plainlyasking
plainlyasking.com
“Ethereal Relaxation” Kevin MacLeod (incompetech.com)
Licensed under Creative Commons: By Attribution 4.0
https://creativecommons.org/licenses/by/4.0/
Source: https://incompetech.com/music/royalty-free/index.html?isrc=USUAN2100031
Changes: Excerpted, faded, volume-adjusted and ducked under narration.
Put yesterday's dinner inside, press a button, and a little later it's steaming. No flame, no glowing grill. So what carried the energy into the food?
This episode follows the energy from the plug to your plate: how a magnetron turns electrical energy into microwaves, why a changing electric field tugs on water molecules, why it isn't a special “resonant” note, why the edge can be hot while the middle stays cold, what the turntable, stirring and standing time actually do, why you can see through the metal mesh in the door, and why heating food doesn't make it radioactive.
0:00 How does a microwave make things hot?
0:19 Electricity becomes microwaves
0:44 The electric field changes direction
1:09 Water has an uneven charge
1:32 Absorbed energy becomes thermal energy
1:59 Food is a mixture
2:26 Not simply inside out
2:48 Reflections make an uneven pattern
3:14 Turn. Stir. Let it stand.
3:42 A shield with a window
4:14 Heating doesn’t make food radioactive
4:39 Follow the energy
Instagram, X and TikTok: @plainlyasking
plainlyasking.com
“Clean Soul” Kevin MacLeod (incompetech.com)
Licensed under Creative Commons: By Attribution 4.0
https://creativecommons.org/licenses/by/4.0/
Source: https://incompetech.com/music/royalty-free/index.html?isrc=USUAN1300033
Changes: Excerpted, faded, volume-adjusted and ducked under narration.
You put a drink in the fridge, close the door and come back to find it chilled. Somewhere around the outside of the fridge, though, warmth is being released into your kitchen. That isn't an unfortunate side effect of making cold. It's the other half of the job.
This episode follows the heat around a fridge's sealed loop: why the refrigerant boils on the cold side, how the compressor makes it hotter than the kitchen, where the condenser sends that energy, why a narrow restriction drops the pressure, how the energy accounting balances, and why leaving the door open can't cool the whole room.
0:00 How does a fridge make things cold?
0:33 Move energy out
1:11 One sealed loop
1:47 Boiling on the cold side
2:22 Work changes the conditions
2:55 Give energy to the room
3:29 Drop the pressure
4:11 Follow one whole circuit
4:48 Energy must balance
5:22 The open-door trap
5:58 Put the hot side elsewhere
6:34 Cold inside, warmth outside
Instagram, X and TikTok: @plainlyasking
plainlyasking.com
“Ever Mindful” Kevin MacLeod (incompetech.com)
Licensed under Creative Commons: By Attribution 4.0
https://creativecommons.org/licenses/by/4.0/
Source: https://incompetech.com/music/royalty-free/index.html?isrc=USUAN1700033
Changes: Excerpted, faded, volume-adjusted and ducked under narration.
A customer writes: the tracking says delivered, but there is nothing outside my door. Before anyone writes a reply, there are other jobs. Which team should see it, what needs checking, and is the system allowed to do anything beyond suggesting an answer?
This episode follows one fictional parcel problem to separate finding information, making a prediction, choosing an action and writing a response. We look at rules and classification, typed answers with a fixed shape, Jev from TypeSafe AI as one current decision-model example, what an 80% prediction really means, how costs turn a probability into a choice, why options and awkward test cases matter, why input is not authority, and who should have the final say.
0:00 How does AI choose?
0:48 From words to choices
1:36 Some choices are rules
2:21 Classification
3:10 A fixed answer shape
3:52 Jev from TypeSafe AI
4:43 A different output interface
5:31 More than one tool can fit
6:22 Give the choice evidence
7:10 What does 80% mean?
8:00 Confidence needs checking
8:47 The cost of a wrong choice
9:46 People set the objective
10:36 Both can be true
11:23 Test the awkward cases
12:09 Input is not authority
12:56 Choosing is not acting
13:41 Does every choice teach it?
14:30 One problem, several tools
15:20 Who has the final say?
Instagram, X and TikTok: @plainlyasking
plainlyasking.com
“Ever Mindful” Kevin MacLeod (incompetech.com)
Licensed under Creative Commons: By Attribution 4.0
https://creativecommons.org/licenses/by/4.0/
Source: https://incompetech.com/music/royalty-free/index.html?isrc=USUAN1700033
Changes: Excerpted, faded, volume-adjusted and ducked under narration.
Ask a chatbot to explain a bicycle to a pirate and an answer appears in seconds. Nobody wrote that exact answer down in advance, so was it found, remembered or made right then?
This episode follows one answer from your question to the words on the screen, no coding needed. We look at how training on examples adjusts a model’s settings, why a raw text predictor is not automatically a helpful assistant, the difference between training and inference, how text becomes tokens and lists of numbers, how context and attention tell one “bank” from another, how a reply grows a piece at a time, why fluent does not mean factual, and how retrieval can bring in current information without making mistakes disappear.
0:00 Where does an answer come from?
0:41 A familiar route
1:46 AI is an umbrella
2:09 Learning from examples
3:02 Parameters: adjustable numbers
3:47 Patterns can combine
4:37 From continuation to assistant
5:28 Training is not inference
6:17 Language arrives in pieces
7:03 A vector is a list of numbers
7:51 Relationships in many directions
8:42 Context changes the representation
9:38 Attention mixes available information
10:39 A distribution over next tokens
11:31 Append, update, repeat
12:24 Context: what is available now
13:14 Fluent does not mean factual
14:05 Find, supply, answer
14:55 Search can match meaning
15:45 A source still needs checking
16:32 The journey of an answer
17:20 The next question
Instagram, X and TikTok: @plainlyasking
plainlyasking.com
“Ethereal Relaxation” Kevin MacLeod (incompetech.com)
Licensed under Creative Commons: By Attribution 4.0
https://creativecommons.org/licenses/by/4.0/
Source: https://incompetech.com/music/royalty-free/index.html?isrc=USUAN2100031
Changes: Excerpted, faded, volume-adjusted and ducked under narration.
A river can run through a freshwater lake, keep travelling, and end up in a salty sea. It is carrying the same water downstream, so why is the ocean salty when most lakes are not?
This episode follows dissolved minerals from rain weathering rocks into rivers, then into lakes and the sea. We look at why evaporation leaves most salts behind in the ocean, why most lakes stay fresh when water flows out, why closed lakes in dry climates can become saltier than the sea, why the ocean is not equally salty everywhere, and why it is not simply getting saltier forever.
0:00 Why is the ocean salty?
0:21 Where the salt comes from
1:04 Evaporation leaves the salt behind
1:54 How salty is seawater?
2:15 Why most lakes stay fresh
2:57 Lakes saltier than the sea
3:19 The ocean is not equally salty
3:39 Salt also leaves the ocean
4:06 Two bowls
4:28 Follow the salt
Instagram, X and TikTok: @plainlyasking
plainlyasking.com
Music: “Moonlight Beach” by Kevin MacLeod (incompetech.com)
Licensed under Creative Commons Attribution 4.0: creativecommons.org/licenses
Source: incompetech.com/music/royalty-free/index.html
Changes: Trimmed, faded, and mixed beneath narration.
You record a quick message, press play and hear someone who sounds almost like you. The words are yours and the accent is yours, but something about the voice seems wrong.
This episode follows the two ways your own voice reaches your ears: through the air, and through the tissues and bones of your head. We look at why playback changes that mix, how the voice you hear while speaking becomes your familiar reference, what familiarity does to your reaction, how microphones, rooms and speakers shape a recording, whether a recording is what everybody else hears, and how to make a fairer comparison.
0:00 Why your recorded voice sounds wrong
0:19 Air conduction
1:01 Bone conduction: the internal route
1:47 Playback changes the mix
2:37 The voice you are used to
3:00 Familiarity
3:24 Microphones, rooms and speakers
3:48 Is the recording what others hear?
4:10 A fairer comparison
4:35 Same voice, different route
Instagram, X and TikTok: @plainlyasking
plainlyasking.com
Music: “Vibing Over Venus” by Kevin MacLeod (incompetech.com)
Licensed under Creative Commons Attribution 4.0: creativecommons.org/licenses
Source: incompetech.com/music/royalty-free/index.html
Changes: A continuous excerpt, entrance/exit fades, volume adjustment and ducking beneath narration.
A QR code with a smudge across it can still return the same message. The missing squares have not become visible again — the code was built with enough extra information for the reader to work out what was there.
This episode looks at how QR codes find their grid, how bits are grouped into codewords, why redundancy is not simply a second copy, and how Reed-Solomon error correction recovers damaged data within limits. We cover the L, M, Q and H recovery levels, why the reader must locate the symbol before it can repair anything, why damage across interleaved blocks matters, and how a small logo spends the same allowance needed for dirt and poor printing.
0:00 A smudged QR code
0:24 Find the grid
0:45 Codewords
1:09 Redundancy
1:34 A missing-number puzzle
1:55 A wrong value
2:24 Reed-Solomon
2:53 Protection takes space
3:17 L, M, Q and H
3:42 Read before repairing
4:07 Block by block
4:30 Logos
4:52 Wrap-up
Instagram, X and TikTok: @plainlyasking
plainlyasking.com
Music: “Wholesome” by Kevin MacLeod (incompetech.com)
Licensed under Creative Commons Attribution 4.0: https://creativecommons.org/licenses/by/4.0/
Source: https://incompetech.com/music/royalty-free/index.html?isrc=USUAN1900022
Changes: Trimmed, faded, and mixed beneath narration.
Imagine driving somewhere, throwing away the engine halfway there, and buying another car for the next trip. For much of spaceflight, discarding very expensive hardware has been part of the plan.
SpaceX wants to change that with Starship and Super Heavy: launch an enormous two-stage rocket, send the upper stage into space, and bring both stages home to be caught by a tower. This episode asks what has to work for that idea to change the price of reaching orbit — orbital mechanics, re-entry heating, thermal protection, the belly-flop, turnaround, economics, and what “reuse” would have to prove.
0:00 Intro
0:59 Catching both stages
1:51 A short history of reuse
4:24 Up versus around
6:06 Staging
7:11 The booster catch
9:16 Falling out of orbit
10:16 Re-entry heating
11:22 Thermal protection
14:48 The belly-flop
15:52 Turnaround
16:56 The economics
17:59 What comes next
20:09 Human flight
21:15 The lifecycle
24:24 Conclusion
Instagram, X and TikTok: @plainlyasking
plainlyasking.com
Music: “Fluidscape” by Kevin MacLeod (incompetech.com)
Licensed under Creative Commons Attribution 4.0: https://creativecommons.org/licenses/by/4.0/
Source: https://incompetech.com/music/royalty-free/index.html?isrc=USUAN1100393
Changes: Trimmed, faded, and mixed beneath narration.
An empty bottle looks like a simple job: collect it, melt it, make something new. So why can’t we do that with every wrapper, tub and broken toy?
Follow a plastic bottle through collection, sorting, washing and mechanical recycling. We explore why resin codes are only clues, why quality and buyers matter, and why layered packaging, thermosets and chemical recovery make the picture more complicated.
Follow your local collection guidance: a material being recyclable somewhere does not mean your local service accepts the whole item.
0:00 Why can’t we recycle all plastic?
0:24 Plastic is a family of materials
0:44 Collection needs a working system
1:04 The number is only a clue
1:23 Sorting suitable materials
1:51 Preparing flakes and pellets
2:20 Mechanical recycling
2:29 Melting is not a reset button
2:39 A useful next job
3:10 Damage to polymer chains
3:34 Layered wrappers and blends
4:05 Why thermosets do not simply melt
4:24 Could chemistry solve it?
4:51 Material, system and next job
Instagram, X and TikTok: @plainlyasking
plainlyasking.com
Music: “Half Mystery” by Kevin MacLeod (incompetech.com)
Licensed under Creative Commons Attribution 4.0: https://creativecommons.org/licenses/by/4.0/
Source: https://incompetech.com/music/royalty-free/index.html?isrc=USUAN1900009
Changes: Trimmed, faded, and mixed beneath narration.
From the publisher's feed