Quinn Batten

here to repost tweets about urbanism, bikes, and city govt in Philly/NYC

[BRV @willrayvalentin]

Hope this helps

[Quinn Batten @quinnbatten]

@archived_videos Of course the LLM summary may miss super important parts of the book! But people tend to heavily discount how likely they are to miss things, make mistakes, and so on.

2/2

[Quinn Batten @quinnbatten]

@archived_videos My guess is readers want the-info-in-the-book and are doing the lowest-friction-thing to get it.
Buy: costs money
Pirate: annoying
Ask LLM: zero friction. Type "summarize" plus book name in chatbox of app you already have. No acquisition cost + lower reading effort.

1/

[Quinn Batten @quinnbatten]

Rule number one: read Vicki's posts as soon as they drop!

[vicki @vboykis]

new post: how I develop recently using local models. the tooling is now good enough to do agentic workflows and everyone should give them a try!

https://t.co/3Tx3CMsNG3

[Quinn Batten @quinnbatten]

Next step in the rankings listicles thing is third-party domains that are controlled by the company that's ranking highest. Yes, that was already a thing before AI, but it was less effective.

3/3

[Quinn Batten @quinnbatten]

Correct. AI lovesss rankings & doesn't (yet) exercise judgment about who has written them.

You'll also see way more spammy reddit/social posts nowadays. Same thing - a human reader was likely to catch onto the spamminess, but AIs aren't discerning about that.
1/

[nxthompson @nxthompson]

This is wild—and likely a sign of things to come as we transition to a web that is optimized for bots more than humans. https://t.co/5sb1cFlBQO

[Quinn Batten @quinnbatten]

So it's more valuable to spam Reddit now than it used to be. And, double-whammy, it's also cheaper to create them now bc you can use AI to generate them. Supply & demand curves both shifted to the right.
2/

[Quinn Batten @quinnbatten]

Lets gooo!!

[» teej @teej_m]

I’ve been working on this at @linear for the last few months and I am stoked to go live today

[Citrini @citrini]

I give it a year until we see a new breed of AI native private equity firms that acquire companies just so they can move their workflows from Claude to open source Chinese models and flip them.

[François Chollet @fchollet]

Some considerations that many folks seem not to get:

1. It can be a bubble even if the tech works. (For instance, if the tech doesn't have a high-demand use case.)

2. It can be a bubble even if the tech works and has strong product-market fit. (For instance, if the tech cannot be economically viable.)

3. It can be a bubble even if the tech works, has strong product-market fit, and has a path to eventual economic viability. (For instance, if profitability takes too long to achieve or makes margin/competition assumptions that fail to materialize.)

4. It can be a bubble even if the tech works, has strong product-market fit, and is currently highly profitable. (For instance, if demand has a hard ceiling and growth stops once the ceiling is reached.)

5. It can be a bubble even if the tech works, has strong product-market fit, is currently highly profitable, and has unlimited future demand.

Literally all it takes for something to be a bubble is for lots of people to over-enthusiastically bet their money on it, and subsequently get panicky.

Importantly, bubbles can be attached both to things that are completely hogwash, like the Metaverse, and to world-changing developments like the Internet or railways. Bubbles don't care. They're brought into existence by the thoughts and feelings of investors, not by actual tech or products.

"The bubble has burst" doesn't mean "the tech didn't work" or "people stopped using the tech." It only means that people got panicky, investor money dried up, and valuations collapsed. Internet adoption didn't stop in 2000.

[Quinn Batten @quinnbatten]

Your phone is blowing up with on-call alerts. Some security issue. You roll out of bed, eyelids heavy. Log on and tap into the war room Teams call. Claude won't do *anything*.

"Chat paused: This topic was flagged." Everyone's freaking out.

Lol.

[John Scott-Railton @jsrailton]

NEW: malware developers added nuclear & biological weapons text to to their spyware.

Goal? To trigger LLM safety refusals... so that their spyware wouldn't be analyzed by an AI security scanner.

Cleanest practical example I can think of for why over-indexing on first order safety alignment is risky.

When closed (and open) models ship with aggressive refusals, they will be sprinkled with second-order blindspots that attackers will discover...and exploit.

We are only in the earliest days of attackers leveraging these features, and it wouldn't surprise me if users systems that need to handle complex cybersecurity issues demand that models be less safety-blunted.

In the weeds: @SocketSecurity's post also shows why intention matters in how you design a malware analysis pipeline to avoid prompt manipulation.

H/T to colleagues that shared this with me https://t.co/f3Aj9TYxU4

[Quinn Batten @quinnbatten]

@zarazhangrui OH this is a great read

[Quinn Batten @quinnbatten]

@TheStalwart "While we adhere strictly to Editorial Integrity..." in the last screenshot 😭

[Quinn Batten @quinnbatten]

@AlexNoonan6 well said

[∿ @somewheresy]

chicago from first principles

[Everything Price Sufferer (but especially eggs) @agraybee]

From 1966 to 2025 we dropped sterile flies over South America that ate screwworm and thus prevented them from spreading, but the le epic efficient cracked coders at DOGE thought this was a silly waste of the ~0 dollars it cost us.

[Arun S. Chopra CFA CMT🧐 @FusionptCapital]

MORGAN STANLEY TO ALLOW HOUSE PETS TO BUY SPACE X IPO SAYS OPEN ACCOUNT IN PETS NAME FUND ACCOUNT AND BUY SHARES 50 DOLLAR MINIMUM EXCEPT FOR RETRIEVERS RETRIEVERS MUST HAVE MINIMUM 1K DOLLARS

[Quinn Batten @quinnbatten]

@MatznerJon Bossshhhh

[Daniel Tenreiro @TenreiroDaniel]

@Pontifex Holy Father the post will perform better if you append the url in a threaded reply as opposed to linking it in the tweet

[Quinn Batten @quinnbatten]

@JaredSleeper The discrepancy btwn Databricks and Snowflake was surprising to me at first. But I actually can think of two AI/ML teams that have had a choice between the two and have much preferred Databricks.