Test dataleaked intotraining setRequirementschanged…againWho lefttrainingrunning ALLweekend?Just addONE morefeaturebeforelaunchJupyterkernel diedmysteriouslyOOMkilled atepoch 99of 100Can weget 99.9%accuracy?Forgot toshuffle thedatasetThe labelsarecompletelywrongSSHconnectiondied mid-experimentForgot tokill myprocess…sorry team!We needmoredataRestartkernel,run all,prayAs permy lastemail…Worksperfectlyon mymachineThebusinessneed has…evolvedActually,let’s pivotthe entireapproachServerrebootedovernight.RIP myresults.A meetingcouldhave beenan emailInternetdied atdemotimeThisnotebookis 2000cells longThisshould besimple,right?Whoannotatedthis?!“Quick 5-min sync”= 45minutesThe serverissuspiciouslyslow todaySomeonehoggedall GPUsCUDAout ofmemoryPetphotobombedthe video callNicepajamatop on thecallExtremeclassimbalancestrikes againStakeholderghostedafter week2Servercrashedduring thelive demoI’lldocumentthis laterWait…which celldid I runlast?Loss wentto NaN onstep 1Groundtruth isn’tactuallytrueCan wemake itreal-time?When willthe model beproduction-ready?Datasethas 47samples.Total.The datais toonoisy touseFree!Test dataleaked intotraining setRequirementschanged…againWho lefttrainingrunning ALLweekend?Just addONE morefeaturebeforelaunchJupyterkernel diedmysteriouslyOOMkilled atepoch 99of 100Can weget 99.9%accuracy?Forgot toshuffle thedatasetThe labelsarecompletelywrongSSHconnectiondied mid-experimentForgot tokill myprocess…sorry team!We needmoredataRestartkernel,run all,prayAs permy lastemail…Worksperfectlyon mymachineThebusinessneed has…evolvedActually,let’s pivotthe entireapproachServerrebootedovernight.RIP myresults.A meetingcouldhave beenan emailInternetdied atdemotimeThisnotebookis 2000cells longThisshould besimple,right?Whoannotatedthis?!“Quick 5-min sync”= 45minutesThe serverissuspiciouslyslow todaySomeonehoggedall GPUsCUDAout ofmemoryPetphotobombedthe video callNicepajamatop on thecallExtremeclassimbalancestrikes againStakeholderghostedafter week2Servercrashedduring thelive demoI’lldocumentthis laterWait…which celldid I runlast?Loss wentto NaN onstep 1Groundtruth isn’tactuallytrueCan wemake itreal-time?When willthe model beproduction-ready?Datasethas 47samples.Total.The datais toonoisy touseFree!

AI lab BINGO - Call List

(Print) Use this randomly generated list as your call list when playing the game. There is no need to say the BINGO column name. Place some kind of mark (like an X, a checkmark, a dot, tally mark, etc) on each cell as you announce it, to keep track. You can also cut out each item, place them in a bag and pull words from the bag.


1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
  1. Test data leaked into training set
  2. Requirements changed… again
  3. Who left training running ALL weekend?
  4. Just add ONE more feature before launch
  5. Jupyter kernel died mysteriously
  6. OOM killed at epoch 99 of 100
  7. Can we get 99.9% accuracy?
  8. Forgot to shuffle the dataset
  9. The labels are completely wrong
  10. SSH connection died mid-experiment
  11. Forgot to kill my process… sorry team!
  12. We need more data
  13. Restart kernel, run all, pray
  14. As per my last email…
  15. Works perfectly on my machine
  16. The business need has… evolved
  17. Actually, let’s pivot the entire approach
  18. Server rebooted overnight. RIP my results.
  19. A meeting could have been an email
  20. Internet died at demo time
  21. This notebook is 2000 cells long
  22. This should be simple, right?
  23. Who annotated this?!
  24. “Quick 5-min sync” = 45 minutes
  25. The server is suspiciously slow today
  26. Someone hogged all GPUs
  27. CUDA out of memory
  28. Pet photobombed the video call
  29. Nice pajama top on the call
  30. Extreme class imbalance strikes again
  31. Stakeholder ghosted after week 2
  32. Server crashed during the live demo
  33. I’ll document this later
  34. Wait… which cell did I run last?
  35. Loss went to NaN on step 1
  36. Ground truth isn’t actually true
  37. Can we make it real-time?
  38. When will the model be production-ready?
  39. Dataset has 47 samples. Total.
  40. The data is too noisy to use
  41. Free!