Cason Cherry

father, husband, #CompSci, #AI, #OMSCS @gtcomputing | ☁️ builder & #engineer @LibertyMutual | avid reader, & maker of great scrambled eggs!

[Cason Cherry @CasonCherry]

@uneducatedecon1 @x_Aurelion @histories_arch Yeah I was going to say there are inherent problems with any monetary system. The Gold Standard was abandoned because of limitations placed on central banks. Commodity systems create rigidity (cacao, gold, crypto) all create deflation. Fiat systems generally create inflation

[Cason Cherry @CasonCherry]

If you chat with the LLM wrapper (Caia) I put on there, let me know. It's basic, but it will do the job.

[Cason Cherry @CasonCherry]

Ok, so I made some long overdue changes, and then went to town with Claude.

https://t.co/O3TiMkO4fA

[Cason Cherry @CasonCherry]

So much about this hurts. My favorite text showing up into publications written with the help of (by?) LLMs is “sure, I’d love to help you write your paper. How about…”

[Jeremy Nguyen ✍🏼 🚢 @JeremyNguyenPhD]

Are medical studies being written with ChatGPT?

Well, we all know ChatGPT overuses the word "delve".

Look below at how often the word 'delve' is used in papers on PubMed (2023 was the first full year of ChatGPT).

[Cason Cherry @CasonCherry]

@tdietterich I feel like this also falls under the realm of slots/intents identification.

[Cason Cherry @CasonCherry]

@tdietterich Good in theory, but there is a number of ethical and geopolitical hurdles that make the conditions here hard in practice.

[Cason Cherry @CasonCherry]

@neilquinn @GeorgeMayer @uddupa People are complicated.

[Cason Cherry @CasonCherry]

Airbender

[Austen Allred @Austen]

What's the right job title for a startup when you want a super scrappy person who can basically do whatever is asked, is self-sufficient, and learns quickly?

But spans multiple departments/roles.

Asking for a friend.

[Cason Cherry @CasonCherry]

@GeorgeMayer @uddupa @neilquinn This is in the eye of the beholder. Some great engineers are awful to manage. Most people just need the right environment to not be “bad people”. Many managers should simply not be managers. The most valuable people can see who should and who should not be leading anything

[Cason Cherry @CasonCherry]

@saiphcita Congratulations on your citizenship Dr. Savage! And thanks for everything you do!

[Cason Cherry @CasonCherry]

Tomorrow: Elon controls you with Neuralink 😂

[Cason Cherry @CasonCherry]

😂 this thread

[Microsoft Teams @MicrosoftTeams]

@hookana We love that 💜 What's your status today?

[Cason Cherry @CasonCherry]

On reflection, my @Docker t-shirt was my all-time favorite swag tee. I was legit sad when it finally wore out.

[Cason Cherry @CasonCherry]

The truth hurts, but every word of this resonates with me. The tech remains largely experimental in most companies with only a few revenue optimizing “applications” to help people do things faster. It not going to blow any bottom lines up except the providers. You know who I mean

[Ed Zitron @edzitron]

Today's newsletter: I believe we're reaching peak AI, and that the industry is going to crash to Earth as enterprises and consumers realize the limits and lack of utility of ChatGPT and co.

And it's all thanks to Sam Altman's empty hype cycle.

https://t.co/gB5eYvQBXm

[Cason Cherry @CasonCherry]

@rdbatch02 Also, I like the monitor in the background, Ryan. This is dedication.

[Cason Cherry @CasonCherry]

@rdbatch02 I shed feedback to Microsoft this morning, “why not just add a flag so people can choose to have the delimiter or not?”

MSFT: 🤦 “that’s a really good idea actually”

Tonight I celebrate.

[Cason Cherry @CasonCherry]

@rdbatch02 This was my life last night when I realized the Azure Pipelines degradations I was seeing were the result of the new Azure CLI rolling out a breaking change that removed the question mark delimiter from their storage tokens. Apparently the world is split where this belongs

[Cason Cherry @CasonCherry]

[2402.06664] LLM Agents can Autonomously Hack Websites

…LLM agents can autonomously hack websites, performing tasks as complex as blind database schema extraction and SQL injections without human feedback. agent [need not] know vulnerability beforehand https://t.co/WgomhQEXnh