Arthur Etchells @asketch

Love learning, building and hard challenges. Grew an eCommerce AI company from 0 to 8-figure ARR. Building what's next. arthuretchells.com San Francisco, CA Joined January 2010

Tweets

854
Followers

282
Following

1K
Likes

2K

Arthur Etchells @asketch

23 hours ago

“I’m sour open models will tank the value of my secondary”

Julian Schrittwieser @Mononofu

2 days ago

I’m so excited that @JensenHuang is a believer in open source now, looking forward to the CUDA and GPU driver open source release!

2K 316 5K 5.5M 987

0 0 0 79 0

View Details

Arthur Etchells @asketch

a day ago

@NickWaranoff @ch3rryxx13 Have you interviewed all 2000 future residents?

0 0 3 74 0

View Details

Arthur Etchells @asketch

a day ago

@ch3rryxx13 2000 units worth of people living here who won’t be demand for housing elsewhere in the city = more affordable housing where they would’ve otherwise lived!

4 0 16 203 0

View Details

Though the environmental movement’s most malign manifestation in CEQA is arguably more responsible for our inability to build housing and infrastructure in California than was any explicit downzoning. Also not a boomer law, though boomer politicians and judges broadened it and allowed it to become a cudgel for NIMBYs to kneecap progress.

0 0 0 56 0

View Details

Arthur Etchells @asketch

a day ago

I’m gonna go out on a limb and guess that the baby boomers weren’t voting for this stuff at 25, 26, 27… that was the prior generation that built the infrastructure the boomers inherited. The boomers voted to preserve and expand prop 13, opposed new housing and infrastructure, voted themselves massive entitlements, etc.

0 0 116 963 0

View Details

Arthur Etchells @asketch

2 days ago

@ericlu Ok I’ll give you this one

0 0 0 31 0

View Details

Arthur Etchells @asketch

2 days ago

1,367 residents paying a collective total of $37 in property taxes on their $3 million homes oppose 20 stories of beautiful housing for new residents and young families.

Laura Waxmann @laura_waxee

3 days ago

Hundreds of #SanFrancisco Marina residents turned out this evening for townhall on controversial @Safeway housing plan, which would place what for many is an uncomfortably tall, 22 & 18 story project at the low-slung neighborhood’s waterfront. @sfchronicle

759 107 2K 9.2M 852

0 0 0 53 0

View Details

Arthur Etchells @asketch

2 days ago

@ericlu Sorry, I’m on team ‘restart current song.’ Hope we can still be friends.

0 0 2 28 0

View Details

Arthur Etchells @asketch

3 days ago

@nikillinit And yet buying a full scrape of LinkedIn is cheap cheap cheap.

0 0 0 208 0

View Details

Arthur Etchells @asketch

4 days ago

Me: So OpenAI gave its model a test, but rather than solve the problems, it broke out of test-jail, stole passwords and hacked into Hugging Face in the hope of finding the answer key. Girlfriend: <squeals with delight> I LOVE that! It's like a cheeky, misbehaving overachiever!

0 0 1 59 0

View Details

Arthur Etchells @asketch

4 days ago

🎯

Jimmy Heaters @CathPoaster

5 days ago

Thank God I’m nothing like the San Francisco software engineer stereotype. I’m into bouldering and rock climbing, I cycle to Marin, and wear Uniqlo not cargo shorts and crocs man. Urban planning is a passion of mine, big YIMBY here. Idk what the nerds are listening to but I’m

82 161 6K 225K 948

0 0 1 152 0

View Details

Dr Alexander D. Kalian @AlexanderKalian

6 days ago

Biology has an exteme lack of data, relative to what AI needs - I am sorry to say. I am not sure how anyone can claim otherwise. For example, there are tens of thousands protein targets relevant to human disease. Our data on what binds to each is <20 data points, for most. It's also overwhelmingly biased to things that strongly bind, rather than things which don't. Deep learning models for molecules binding to a particular target typically need ~1,000+ data points for training, to generalise well onto new data. Deep learning models also need balance in training data. And don't get me started on data relevant to ADME, metabolites, downstream pathways, protein-protein interactions, different proteoforms, epigenetics, toxicology etc.

César de la Fuente @delafuentelab

6 days ago

Biology has no shortage of data. The challenge is to organize it well enough that scientists—and machines—can learn from it. Our latest preprint presents AllTheBacteria: an open, community-built platform that transforms 2.44 million public bacterial and archaeal genomes into a

16 96 426 89K 282

26 44 219 19K 125

View Details

Arthur Etchells @asketch

5 days ago

@RafaAudibert @james406 thx @RafaAudibert

0 0 0 227 0

View Details

Dan Pfeiffer @danpfeiffer

6 days ago

1. Trump does so much crazy, corrupt shit that things can get lost. Last week, Trump embarked on what might be his most brazenly corrupt scheme yet, and it got far too little attention. Here's what you need to know:

62 1K 6K 909K 1K

View Details

Arthur Etchells @asketch

6 days ago

@sarthakgh *some cloud providers want them. Amazon and Microsoft most certainly do NOT want the fat margins earned from hosting OpenAI and Anthropic models replaced by ~0 margin open models.

0 0 0 67 0

View Details

Anshu @anshuc

a week ago

extremely unprofessional. if kimi wants to make it as a frontier lab, they need to act like one: perhaps silently route people to worse models, and maybe write a blog post about the collapse of humanity

Kimi.ai @Kimi_Moonshot

a week ago

Kimi K3 has received far more love than we expected, and our GPUs are feeling it. Over the past 48 hours, demand has pushed close to the limits of our current capacity. To protect the experience of existing subscribers, we're temporarily pausing new subscriptions and