---
title: "Slides: ZSQb5fzRFPw"
category: "slides"
video_id: "ZSQb5fzRFPw"
sourceLabels: ["Public YouTube video frames", "Public YouTube metadata"]
---

# Slides: ZSQb5fzRFPw

## Source Video
[](https://www.youtube.com/watch?v=ZSQb5fzRFPw)

## 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
![[assets/slides/ZSQb5fzRFPw/slide-001.jpg]]

OCR text:

> LAB & PLATINUM SPONSORS
> Amazon AGI Lab ANTHROP\C Google DeepMind MINIMAX
> OpenAl ZB ss akamar arize 2YS %Braintrust bright data
> B Srowserbase %docker -NEO4) ORAcLeE PayPal ¢&
> -y reducto (yy Sonar’ © togetherai (ff Unblocked €} WorkOS

![[assets/slides/ZSQb5fzRFPw/slide-002.jpg]]

OCR text:

> Al Engineer
> World's Fair

![[assets/slides/ZSQb5fzRFPw/slide-003.jpg]]

OCR text:

> Cua
> World's Fair AlEngineer

![[assets/slides/ZSQb5fzRFPw/slide-004.jpg]]

OCR text:

> ran Tne pUaxen]
> World's Fair

![[assets/slides/ZSQb5fzRFPw/slide-005.jpg]]

OCR text:

> eval kd alg |
> RO Mem ICRA LEIETOR RCO)
> Se ec me
> eee Engineering the future of Al

![[assets/slides/ZSQb5fzRFPw/slide-006.jpg]]

OCR text:

> World's Fair Al Enginee A GUI-first operator Cua
> The old loop wired the model straight to the GUr prompt and screenshot in action out.
> MOCEL
> reapossetcltot.nesponses.createi ecde.-conpter-use.geeviee iapoteI tootarll "content' "role"truser" "isptoy.aidth":12s, "sisplay.helght:7e, tyae':coapotorose.orev1es (type*"ingot_text'text':tosk) yengu.gmgr：r） pro tex cor TOCC Opon INPUT Thok
> Thon
> cus.ai CUTPO
> World'sFair Computer Use TRACK 7·JULY 1.2026

![[assets/slides/ZSQb5fzRFPw/slide-007.jpg]]

OCR text:

> Worid's Fair Cua Aa SkyLight Recipe
> Focus without roise. post to pid primo Chrom um.
> SLE
> S
> Andk LRK
> live
> Pnote
> Workd's Fair ComputerUse TraCk70JulYt.2026

![[assets/slides/ZSQb5fzRFPw/slide-008.jpg]]

OCR text:

> Al Engineer ;
> ela kta | ae
> One interlace to rule them all.
> HH 13
> Windows TATOAN
> TRACK 7 * SULY 1, 2026
> ee Computer Use

![[assets/slides/ZSQb5fzRFPw/slide-009.jpg]]

OCR text:

> Tern oe | |
> Now: can the agent usc the computer?
> a Microsoft
> CoE Daniele Cua Bench Dillon DuPont
> ee Computer Use

![[assets/slides/ZSQb5fzRFPw/slide-010.jpg]]

OCR text:

> rld's Fair Cua One open catalog. 130TASKS-CUA-SENCH
> 130 42 5
> VERIFIABLETASKS ENVIRONMENTS PLATFORMS
> Scb run dataset cua-bench-kicad -m sonnet-4-5
> Linuxndondridmcose cua.ai
> Vorld'sFair Computer Use TRACK 7·JULY 1.2026

![[assets/slides/ZSQb5fzRFPw/slide-011.jpg]]

OCR text:

> pa :
> World's Fair
> 
> Reco —
> OTRO INNS Kood
> trust the nn
> nraeCUP : .
> 
> TRACK 7 *« JULY 1, 2026
> eee Computer Use

![[assets/slides/ZSQb5fzRFPw/slide-012.jpg]]

OCR text:

> pbs ,
> World's Fair |
> Interruptible Generation
> ————7 | i
> | im t= (rt
> a | | te hh aoe, ee
> ba Patera eae ae
> 
> TRACK 7 * JULY 1, 2026
> Computer Use

![[assets/slides/ZSQb5fzRFPw/slide-013.jpg]]

OCR text:

> World'sFair AIEngineer Cua
> BORKER UTILIZATION·GENERATION VS.IOLE
> PRESENTEDBY
> Microsoft
> enerafion
> every 1dte gap is a GPu saiting: nininize it cua.a1
> ComputerUse TRACK7·JULY1,2026

![[assets/slides/ZSQb5fzRFPw/slide-014.jpg]]

OCR text:

> World's Fair Cua
> Keep a warm pool.
> GPUs spend less time waiting on cold starts. Warmenvironmentsscalewithdemand.so
> war poot tracks workload·GPUs keop norking cua.ai
> Computer Use TRACK7·JULY1,2026
> Ating

![[assets/slides/ZSQb5fzRFPw/slide-015.jpg]]

OCR text:

> [.. AlEngineer
> ., ae 2 , ae el — @

![[assets/slides/ZSQb5fzRFPw/slide-016.jpg]]

OCR text:

> PUR IR et a ierare
> World's Fair

![[assets/slides/ZSQb5fzRFPw/slide-017.jpg]]

OCR text:

> World's Fair AEng


## 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.