Edward Krueger

repo: github.com/edkrueger/open…

Have a podcast or other audio file that you want to transcribe and summarize? I wrote this script to transcribe and summarize an interview with Geoffrey Hinton from a podcast. I’ve added some instructions to make it easier to use. Any features that you’d like to see added?

#economics #CFO #advertising #ROAS #metrics #MFA #budgetmisallocation

MFA sites are designed to optimize this second class of metrics in your reports, but they have some of the worst performance (often negative ROAS) when revenue can be attributed. Having a large number of MFA sites in a report is clear evidence of a budget misallocation.💸🚫

Why is MFA suddenly in the industry news? One reason is the economic downturn: Your CFO doesn't care about advertising vanity metrics, especially during an economic crunch. Companies are under pressure to show profitability through increasing revenue or cutting costs. 💰

What is (Made for Advertising) MFA and why should advertisers care? - It’s sites that you know to bounce from - It’s sites that slow your browser even when not the active tab - It’s sites you add to you block lists As an advertiser, you should care, b/c everyone hates these😠🚫

Curious about the history of AI or want to challenge your notions about AI? Listen to this! Geoffrey Hinton gave a great interview to 'On The Media' recently. Starts at 16:55, ends at 25:08 Link to interview: wnycstudios.org/podcasts/otm/e…

Thoughts? If you're interested in discussing potential applications for LLMs or any other technology at your business, send me a DM! 📷 #gpt #ai #llm #bard

Bard wouldn't limit its answers to just a category. Even rewriting the prompt with an example didn't affect this behavior. We could have said that Bard hallucinates 100% of answers if we weren't charitable. Instead, we checked if the outputted text container was one of the labels

What's better, Bard or ChatGPT? 🤔 One way to evaluate a model’s language representation is to see how well the model does at an upstream task. Here we check if Bard or ChatGPT performs better for a zero-shot classification problem. 🧵🧵🧵

Magic the Gathering Arena: Winning with #Probability - How to use probability and #excel to win tournaments Read our latest article on Medium - medium.com/towards-data-s… #DataScience #GameAnalytics #Mathematics #ProbabilityDistribution #StatisticalAnalysis #GameStrategy

We are proud to introduce Peak Values Consulting, a company that combines our expertise in #datascience, #softwareengineering, #AI and #scientificcomputing with SME to provide cutting-edge solutions for your business needs. Follow @peak_values peakvaluesconsulting.com

Last week, I attended the graduation dinner for the #UniversityofTexas MA Economics program's class of 2023. It was great to catch up with students, talk shop and hear from them as they begin their journeys. Congrats, UT #MAEconomics Class of 2023! #UTGrad #Classof2023

Showing ads to only engaged users increases viewability and decreases bid volume without decreasing revenue. Check our guest post at @deepsee_io.

BREAKING: We're pulling back the curtain on "rewarded traffic," and showing you how it's actually delivered. Big publishers have been using this inorganic traffic to fulfill direct campaigns for years. deepsee.io/blog/rewarded-…

9.8% of programmatic display & video ad budgets in the US are being wasted on 'Made For Advertising' inventory. Let that sink in. Full report by @ebiquityglobal will be released tomorrow. Thanks @Ryanbarwick at @MarketingBrew for reporting on this. marketingbrew.com/stories/2022/0…

github.com/ChristopheGarc… crime-ut-data-class.herokuapp.com

Check out this project from my latest cohort of the Data Analytics and Visualization Boot Camp at Texas McCombs School of Business. The group collected data and cleaned data on crime, food scarcity, and weather and merged the results into a dataset.