| theme | background | title | info | class | highlighter | drawings | transition | mdc | fonts | |||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
seriph |
Welcome to Slidev |
## Slidev Starter Template
Presentation slides for developers.
Learn more at [Sli.dev](https://sli.dev)
|
text-center |
shiki |
|
slide-left |
true |
|
Air quotes & acronyms: why trust AI if you can't source it local or see the capital & incentive stacks?
(Especially in health care, where the stakes are high and individuals are not risk-bearing entities... yet! Current incentives: misaligned.)
Pref. KPI: QALYs or DALYs / FLOP to allocate global resources; not in an EA way.
Email: [email protected] | Twitter: @thejaan
Join at menti.com | use code 2871 8158
https://www.menti.com/alz2p6gwkwv3
layout: image-left
image: finnsjo_garden.jpg
-
How cool you think you are vs how cool you are (re: sauna)
-
Cool stuff: anarchist cookbook, Justin Bieber slowed 800x, Nelly Arabic remix
-
Market cap of Microsoft Research vs Wallenberg.com vs venture capital in EU
-
British Broadcasting Corporation & the Cotton Capital - Adam Curtis versus a hip replacement (if done too soon, will cost you and health system more!)?
-
Baby G
-
Data center constraints & electricity grid constraints
https://www.datacentermap.com/
https://www.dutchdatacenters.nl/en/map/
Steve Hayes: "Verbal events are relational operants"
https://arxiv.org/abs/1609.05807 -
Follow-up work:
https://link.springer.com/article/10.1007/s13347-022-00584-6
https://arxiv.org/pdf/2302.06347
https://www.ncbi.nlm.nih.gov/pmc/articles/PMC10148336/
UNDP advising: https://digitalpublicgoods.net/AI-CoP-Discussion-Paper.pdf
opensource.org advising: https://hackmd.io/@opensourceinitiative/osaid-0-0-8
layout: image-left
image: bismarck_2022_family_map.png
"Esse non videri"
(teaser)
Renting to own vs renting to live:
Insulin prices c.a. 2024-06-11
layout: image-left
image: hoang_2023_power_map.png
Spiderweb Capitalism: How Global Elites Exploit Frontier Markets - Kimberly Kay Hoang (Princeton Press
https://poly.cam/capture/7AC16C10-AEAC-46E5-A9CD-280721EA08E3
layout: image-left
image: https://www.payless.health/payless.health-linknyc-campaign.jpg
4000+ hospitals' price sheets: https://data.payless.health/#hospital_price_transparency/
Data model: https://github.com/OHDSI/CommonDataModel/tree/payless_health
AI: http://arxiv.org/abs/1904.05342
20% of GDP: ($5T)
So far helped allocate $1B of capital (story of the $25M bill that resulted from a market maker trying to change insurance products influenced by our open source tools)
-> https://motherduck.com/blog/introducing-column-explorer/
https://motherduckcommunity.slack.com/archives/C059BKPAPC5/p1717864545260129
Hi Jaan:
There's a lot to discuss here, however generally I agree with the possibility of misaligned incentives. Cloud data companies know there a gravitational force to data due to joins, and many of the easy paths bias toward entirely remote data, transforms, compute. Even the asymmetry of AWS egress/ingress costs shows this bias. To be fair this is often efficient, minimizing data transport. However I think you can see the misalignment in the customer pressure to get data lakes as first class peers to internal tables, and in the workflow of ad hoc notebooks, where roundtripping to local is a common practice for it's flexibility.
We think there's a lot of potential to making it much easier to mix and match local and remote sources in one query, and in easier roundtripping. It certainly makes it easier for my ETL jobs to INSERT INTO instead of always dropping into S3, for example.
Let me discuss internally and come up with more specific response with examples. It's a busy week here with our GA launch, but maybe this is worth a call with you to brainstorm a bit more.
Major & minor arcana (Tarot de Marseilles / Adventure Time illustrator Jesse Moynihan) for the NAICS classification codes for industry:
Homework: redo this for EPO data (email me if you don't have free access or need 3.5TB of storage on our AWS S3 bucket :)
Previous example in health care: https://colab.research.google.com/github/onefact/loving-the-baseline/blob/main/nearest-neighbors.ipynb
- Read worksheet on large language models (LLMs) from Ivory Field, experts in generative AI syllabus design.
- Log in to both claude.ai and chatgpt.com and submit one message and ensure you can read the LLM output.
- Pack your laptop for tomorrow.
- Write down the exact time and where you intend complete this homework tonight or tomorrow morning (should take 5 minutes).
Bonus: use an LLM of your choice to copy and paste the Mermaid-syntax code into https://mermaid.live/ and generate a diagram of incentives for a process where you experience maximum frustration or lack of agency (lived experience, personal, systemic, sector-based, acronyms or not, etc). Make sure all cash flows, sponsors, stakeholders are represented and revise the prompt until it captures your lived experience. Ask me or an LLM for help if blocked.
Hover on the bottom-left corner to see the navigation's controls panel, learn more
| right / space | next animation or slide |
| left / shiftspace | previous animation or slide |
| up | previous slide |
| down | next slide |
Here!