Slides: 20 days of compute vs 7 hours: rethinking what state-of-the-art means — Bertrand Charpentier, Pruna
Source Video
20 days of compute vs 7 hours: rethinking what state-of-the-art means — Bertrand Charpentier, Pruna
Relationship To World's Fair 2026
These slides are extracted from a public AI Engineer YouTube video connected to World's Fair 2026. Speaker-matched clips are supporting context unless later confirmed as exact session recordings; official livestream recordings are day-level/event-level source material.
Related Scheduled Sessions
- No individual scheduled session mapping has been assigned yet; treat this as an event livestream deck.
Extracted Slides
OCR text:
What Model is ts
Google DeepMind
| wef,
OCR text:
Use-case: Find the most performant image editing Al model
of
Fe il Step 1: Find a leaderboard ila was Ceomammert fap
cd * ~- Step 2 Select the ==
5g * bd
cad Ort ege ss tae vete
n © Germey 3 re reage Prone 2K aE Pn,
wa © Gare 1 Pro trege Pewee ee
“ © Cron enepre Lent ssa ees
at fescoe Avnet
a
| Al Engineer |
EUROPE
: All
OCR text:
Problem1:Eachpublicleaderboardhasadifferentranking
AIE
(LM)Arena
Artificial Analysis
DesignArena
Engineering the future of Al
OCR text:
ee — -e
a & “Ai ts —_ 8 ie ee
* re hd ee ae wwe _
fier bee Bae -,,
* a wo IDS Tt eee eea
* x aoe teow ve —_ oy oe =
ke Wes Se, Me eG
Pi eer cass oem eke tn ae) 7 a Z 7
fare era ee ee en caer Ror Bi renn men sare)
lana
| Al Engineer |
EUROPE
OCR text:
Use-case: Find the most performant image editing Al model
a . - Step 1a Manua: inspection
" y
* bd
a nad
~- Ld a \ Ty,
rs, of 44
) . rede gy ee
ar. ne
ae (ca eat is Se
Google DeepMind
ef,
OCR text:
i — ~ ae wes aim
t= |- h.
ate i I: I:
ial a SLD SILAS LP SA LL ALPS PL PE LP SL 7
* al PP COOL E SG COOP DELLE LOL EOS OF
* bd
oa
ce
i}
| Al Engineer |
ab elas
a ,
OCR text:
fe I: {2
a i I~ I:
. * NA PELEL SS POL EEA tte PPO SE ELE
* Ld 7
a * ee
; I: E rE:
F ao: oat ual
SS REESEE |: oo REESE te SREEEE
LOSCE, reed “fe CEE, Bnf. yet ALPS, ‘ wy we
' , 5 L 28 RY —
|
2 tJ
Engineering the future of Al
aT ee
OCR text:
What Model is cy
There are multiple SOFA Al medeis
—— —
WH Se Paes ae SE SH ON aSE ei ties SEE BRS RD
re a 7 ce = — oe :
ke meee, _ . =, estink,? : _.
: = _
an
| Google DeepMind
a ee
OCR text:
What Model is ?
There are multiple SOTA Ar moceis ,
Penn
Mare
* S 3
a
|
| AlEngineer |
se) Ke) a3
b, — .
OCR text:
HowToReachSOTAPerformance?
Option 1:Use optimized Performance models
AIE
Video Endpoints
Image Endpoints
TextEndpoints
s
leplicote
aws
Runwar
Segmind
WIRO
Engineering the future of Al
OCR text:
HowToReachSOTAPerformance?
Option 1:Use optimized Performance models
AIE
Video Endpoints
Image Endpoints
TextEndpoints
s
eplicate
aws
Runware
Segmind
qeley
WIRO
Option 2:Develop and apply inference optimization
PrunoAl
AI EfficiencyPackage
Al EfficiencyMaterials
Al Efficiency Courses
mesllibub.comPrunaAllcza
nollaihub.comPnnaAicon
AIEngineer
rinanr
Slide-Derived Subjects To Review
Subject extraction uses video title, related session titles/descriptions, transcript context, and OCR text when available. OCR is best-effort and should be reviewed against the embedded slide images.