lawrebx

[lawrebx @lawrebx]

You won’t know a bubble while you’re in it, but there will be signs…

[Anjney Midha @AnjneyMidha]

AI is Eating Software

(Highly biased math) Across the handful frontier AI startups I’m an angel/board member in, YTD sales have gone from ~$1.3B to ~$8B+

Meanwhile, publics are bleeding

The single most rapid reallocation of value from legacy companies to startups in history?

[lawrebx @lawrebx]

Everyone is blaming AI for new grad unemployment - it’s the economy

[lawrebx @lawrebx]

The analog premium

[Gergely Orosz @GergelyOrosz]

Hate to say it but the future of shopping is not “tell an AI agent what you want and it figures out and buys it for you from a shady site that ships the wrong thing”

It’s “use a trusted app/site with great customer support that has the largest selection of goods and ships today”

[lawrebx @lawrebx]

@tomfriedman https://t.co/HJm8vWEboO

[lawrebx @lawrebx]

Many laypeople misunderstand the underlying nature of AI - at once ascribing unrealistic abilities and understating the risks.

@tomfriedman hits on an important point - AI will be shoehorned into everything - landmines everywhere.

The analog premium will become a big topic.

[lawrebx @lawrebx]

@ergestx So much context is encoded in the necessary feature engineering - I don’t see how you can commoditize it. At some point, the abstraction breaks down and the analyst role become a creative vs technical endeavor.

That’s why most companies struggle to scale beyond Excel.

[lawrebx @lawrebx]

Then what explains their dismal performance for the last 15 years in anything outside of search?

RTO is always RIF in disguise. If a software company can’t work remotely or provide work life balance, it’s a mgmt skill issue.

[lawrebx @lawrebx]

Poor KPI design coupled with high-dimensional decision space

[Jonathan Blow @Jonathan_Blow]

Why does software almost always do the opposite?

[lawrebx @lawrebx]

@Jonathan_Blow Poor KPI design coupled with high-dimensional decision space

[lawrebx @lawrebx]

Leaded gasoline

[LindyMan @PaulSkallas]

Cities used to be filled with tight knit communities in neighborhoods where everyone knew each other. Kids played outside and roamed around and no one cared.

Then mass suburbanization happened in the 60s/70s. American cities became high crime places. Everyone became anonymous. No one knew their neighbor

Serial killers could get away with murdering random people and no one would know or care. Over and over. You could get away with it.

Then in the 90s/00s forensics, cameras, then dna matching got better and serial killers had a risk of being caught by the police. So it stopped.

[lawrebx @lawrebx]

Proof positive that overwhelming talent and expertise is often non-transferable between domains

[Senior PowerPoint Engineer @ryxcommar]

whether Boeing or Crowdstrike, these companies take shortcuts on quality assurance processes because the c-suite sees QA as a cost center that doesn't impact the stock price (until it does). this is so obvious that i'm not convinced any of the morons chanting "DEI" have real jobs

[lawrebx @lawrebx]

Replacing action with litigation is what has stymied American growth in the past couple of decades.

This decision puts that trend into overdrive. WCGW.

[Austen Allred @Austen]

The Supreme Court's Chevron ruling may be most impactful things to happen to startups in a long time, in ways that people don't realize.

A thread:

[lawrebx @lawrebx]

Your boss also needs to be upwardly mobile themselves.

If you make them look too good, but they are in a terminal role, they view you as a threat to their job.

[BuccoCapital Bloke @buccocapital]

One career thing people always get wrong:

You can’t get promoted unless your boss likes you!

It’s true at big and small companies. Your boss has to go to bat for you. That requires political capital

Make them look good. Make their life easier. Simple. People always fuck it up

[lawrebx @lawrebx]

Says more about investor sentiment than anything

[Gergely Orosz @GergelyOrosz]

Tech companies now announcing they are hiring based on merit:

What exactly have you been hiring for, until now?

(Needing to make loud statements on this suggests hiring was based on something else at these companies - unlike the silent majority that always hired based on it)

[lawrebx @lawrebx]

@sagworm Time to build an API…

[lawrebx @lawrebx]

@buccocapital But the meme chats go a long way for morale

[lawrebx @lawrebx]

Kristi Noem’s reaction

[Senior PowerPoint Engineer @ryxcommar]

If you asked a dog "what is 2+2" and the dog answered "3", it's reasonable that your response the first time this happened would be "holy shit a talking dog" and that your response the 100,000th time this happened would be "this talking dog is full of shit."

[lawrebx @lawrebx]

Engineering is magnitudes easier than innovation.

[Arynn does data @ArynnPost]

I struggle to think of new ways to create or use data without it being applied.

And that's probably a good thing?

People build all sorts of apps and systems to collect data. But like...why? My introduction to data was with the premise that you collected it only AFTER you had a research hypothesis or purpose.

So when learning coding in some course, and the prompt is to create an app, my brain just goes "y tho".

If the app is to provide a service to someone, fine. I get that. But if the app is also designed with the data output you can pull from it in mind... why does it feel like people skip a critical planning step?

Just throw all the data in the kitchen sink and figure out what's useful later. Why isn't there a plan of what output will be useful before there is output?

Future you problems don't have to be future you problems (sometimes).