evergreen gambit @evergreengambit

Great analysis. Precisely why Brian Niccol at Starbucks has been so desperate to reintroduce the last-mile. Incredibly difficult to do at scale. Leaves you in this valley of death where you need to continuously grow for your shareholders but risk losing the "last-mile"

illiquid @lefttailguy

a day ago

Chipotle is less consistent and lower quality than the neighborhood taqueria, Starbucks is less consistent and lower quality than the neighborhood cafe. The proverbial La Marzocco espresso machine has democratized "quality" in the generic sense, and therefore the last-mile

5 3 62 3K 38

As much as the app layer looks attractive to play in, the paradigm is different today. App startups are paying inference costs to the same labs they compete with. They're digging with shovels while the labs have bulldozers that are getting exponentially bigger. Ideally, you want the bulldozer maintenance companies (security, compliance, governance, oversight)

@EoinHiggins_ as tempting as it is to say “oh ai make ppl lazy”, these studies fail to capture the percentage of students using ai to master gaps in their knowledge. not using the superintelligent entity with infinite patience to learn is simply just inefficient

@VCBrags maybe 2 weeks ago. "taste is the moat" is already outdated. your moat of finding the optimal serif font for your ai gtm startup website is gone

@Ovahdozz if you're not getting embarrassed, you are not trying hard enough

@s8mb Huge bifurcation between kids using AI to master every gap in their understanding and kids using AI to ignore learning

Wrong. No one knows anything including the economist at Goldman. Unless he was in the offices of OpenAI in 2022 behind Altman's computer there was no way anyone could have known what the future held. Unless you are at the frontier there is no way of knowing what the future of work looks like. You will interact with new form factors and technologies in ways you did not envision to accomplish tasks that did not previously exist.

*Walter Bloomberg @DeItaone

2 days ago

GOLDMAN SACHS: AI COULD TEMPORARILY DISPLACE 9% OF WORKERS Goldman Sachs estimates that a 15% productivity boost from AI could temporarily displace around 9% of workers during the transition. Historical research shows every 1% productivity gain from technology raises

65 62 424 178K 90

The first rule of cybersecurity is to have fun and be yourself

If you think these people are vulnerable, I'm worried for what happens to the rest of the globe that doesn't get to participate in the massive economic event that AI is. We're already starting to see signs of this in Europe as it got left behind in the internet revolution. European capital markets now depend entirely on a few key, protected industries to drive their GDP while the Americans and Chinese have used their tech to unlock new growth vectors. They are now faced with a growing debt balance with no easy way out, while America and China continue to advance at breakneck speed. In a simplistic sense, the events you see in Paris today are all downstream of Europe not having their own Microsoft. Europe's greatest mistake was sitting on their hands in 2000. Hard to imagine that AI won't supercharge this phenomenon even further. Immense economic value will be generated by the countries that lead the frontier and the already growing disparity between the super and middle powers will grow exponentially. Europe will be closer to Brazil than to the US or China.

Rational Aussie @rationalaussie

2 days ago

It's important to remember that most people are still living in the world where they'll be an accountant for 30 years in a computer job and then retire and grow old. At some point over the next few years all of these people will undergo extreme ontological shock as all the

76 115 2K 117K 357

1B series A is crazy

Flapping Airplanes @flappyairplanes

2 days ago

We’ve spent the past year developing new primitives for GPUs. We believe this enables new algorithms to change how machines learn, open new markets, and dramatically reshape AI’s cost structure. We have raised a $1 billion Series A, co-led by Index and Lux.

288 221 4K 824K 2K

Talking to my peers across finance, I'm seeing an increasing amount wake up to the paradigm shift that hit software workers a couple quarters ago. Models, decks, and other aspects of the daily finance grind are getting one-shotted by Claude. This leads to me to question what the future of work looks like in a world that is accelerating faster than ever. It's not a hard jump to assume all knowledge work is done by AI models in the very near future. Every part of the workflow done as it gets exponentially better. At one point it's hard not to imagine a future where one business's agent contacts the standing agent of an M&A shop, which then exchange details, numbers, terms, and other encoded nuances that leads to a marketing deck automatically spun up to send to potential buyers all within the same day. Yes, there might be points where a human intervenes to sign an NDA, hop on a quick call, or check a box but the majority of the process essentially becomes frictionless. We work on devices that our grandparents never envisioned, doing work that didn't exist before. People tend to extrapolate the dynamics of their current livelihood into the future but grossly underestimate how things might operate in a new paradigm. In my prior life as an investment banker, Excel modeling was somewhat of a gatekeeper to the industry. A significant part of your job depended on you being efficient with Excel, building multi-tab models, and figuring out why something returned a #REF error. We are already at the point, where we are at a higher level of abstraction on these tasks and will continue to get higher. At a one point, agents come to run all infrastructure without you seeing it, just like code does now. Want to sell your business? An agent runs 24/7 until your business is sold at the price you set. All you have to do is sign 1 NDA and click 1 button. Want to make truffle tagliatelle for dinner? Agents scan your fridge, understand what ingredients you have/don't have, contacts the warehouse, and ships it to your door by the time you're home. The things we do not have time or effort to optimize for will be optimized. As many people are, I'm starting to zoom out and recalibrate my own career and life as we approach a point where technology goes parabolic, if it already hasn't