Andrew Ho ✓
@andrewho03 · San Francisco, CA · joined 23 Jun 2026
Prev: @OpenAI
18 264Followers
349Following
1 196Posts total
3MViews on collected posts
Derniers posts
Recherche de nouveaux posts…
It’s always interesting to me that people go super hard with fraud to the point that it’s obvious and they get caught. I bet you could get away enriching yourself by a low six figure sum in perpetuity via the exact same method and nobody would ever notice.
5K views · 33 likes
· 0 reposts · 4 replies
Open on X →
I don't know, I think it smells kind of nice.
4.3K views · 11 likes
· 0 reposts · 2 replies
Open on X →
Absolutely one of the worst things about visiting China
149.5K views · 2.8K likes
· 27 reposts · 64 replies
Open on X →
Are there any compelling reasons to buy a Switch 2? The games available seem somewhat mediocre. I'd be interested in language learning purposes though, for example playing video games in Chinese / Japanese / Korean, curious to hear others' experiences.
4.6K views · 20 likes
· 0 reposts · 5 replies
Open on X →
@andrewho03 @AgustinLebron3 Yeah this is pretty fundamental to the whole understanding of where LLMs are placing us.
They are busywork generators. And they fool us into thinking the busywork is useful.
I'm building out a list of films I want to watch right now with Claude and i
429 views · 3 likes
· 0 reposts · 0 replies
Open on X →
@andrewho03 @TheStalwart Yeah, but you're not trying to hype us up for your impending IPO, are you?
12.9K views · 25 likes
· 0 reposts · 1 replies
Open on X →
@andrewho03 @AgustinLebron3 But wasn't this always the case?
Before AI, I remember spending weeks on tasks, with the result ending up completely useless because of change of path/strategy.
At least with AI the wasted time is only hours and not weeks
2.4K views · 21 likes
· 0 reposts · 2 replies
Open on X →
@andrewho03 @AgustinLebron3 I would add that there is also a cost & inference capacity bottleneck
I don't think we can see 50%/100% productivity jumps without enough inference capacity, even if we get AGI/ASI
8.9K views · 58 likes
· 2 reposts · 1 replies
Open on X →
@andrewho03 @AgustinLebron3 > The ability of LLMs to make you waste time doing stuff that is not actually useful is massively underrated imo
Interesting, I can't really think of examples of this! Though I can definitely think of examples of "cadillac tasks" https://t.co/CBw62lB2
6.8K views · 14 likes
· 0 reposts · 1 replies
Open on X →
@AgustinLebron3 The ability of LLMs to make you waste time doing stuff that is not actually useful is massively underrated imo; I spent for example 10s of hours earlier in the year preparing random legal documents with ChatGPT but in retrospect I was too hasty and none of that wo
51.4K views · 585 likes
· 29 reposts · 15 replies
Open on X →
@AgustinLebron3 Not really observations that others before me haven’t made, but:
- Despite the seemingly magical nature of LLMs, reflection over a >3 month timescale suggests my total productivity hasn’t increased by over 100%, or perhaps even by over 50%, and a lot of time is a
369.4K views · 2.8K likes
· 288 reposts · 82 replies
Open on X →
@andrewho03 Can you elaborate? Would be useful, given you're in a fairly unique position.
50.9K views · 144 likes
· 0 reposts · 4 replies
Open on X →
@andrewho03 I do wonder if part of it is that the people who tend to work at labs (at least the ones prior to like 2024) are also the sorts to get the majority of their social interaction *from* work, and so it sort of recursively feeds a doom loop of socialization about the Impo
7.8K views · 30 likes
· 0 reposts · 3 replies
Open on X →
You know, I might be one of the few people in this world who has worked at a frontier lab, observed the nature and pace of research progress, and actually become more bearish on AGI timelines (relative to pre-employment baseline) as a result.
430.4K views · 2.2K likes
· 67 reposts · 52 replies
Open on X →
SITUATION DETECTED: Sam Altman told Axios ahead of the GPT-Astra launch that OpenAI has "much, much more capable models coming soon," and that "no one intellectually honest can look at what's happening and not feel the weight of responsibility in front of us."
626.7K views · 3K likes
· 110 reposts · 41 replies
Open on X →
Today is my last day at @OpenAI. I'm glad to have spent the last eight months of my life working here!
I'm starting a new company focused on the production of high-quality reinforcement learning datasets:
1. The generalization ability of LLMs is clearly very poor, with "spiky"
1.2M views · 5.7K likes
· 291 reposts · 254 replies
Open on X →
Face aux comptes de taille comparable
17 posts des 90 derniers jours, à côté de la tranche de 10K–100K abonnés. diffusé largement, mais peu de ces spectateurs réagissent.
Vues médianes12 863ce compte924médiane pour 10K–100K
Portée, %70.43%ce compte3.62%médiane pour 10K–100K
Engagement, %0.54%ce compte1.52%médiane pour 10K–100K
| Indicateur | Ce compte | Médiane pour 10K–100K | Rapport |
|---|---|---|---|
| Vues médianes par post | 12 863 | 924 | 13.9× |
| Portée (vues ÷ abonnés) | 70.43% | 3.62% | 19.5× |
| Taux d'engagement | 0.54% | 1.52% | 0.36× |
Autres comptes de cette tranche → Comparer avec un autre compte → Comment ces repères sont établis →
Growth & engagement
How the posts we collected actually performed: views and reaction rate post by post, what the audience did with them, and where the follower count goes.
Views per post
7.8K3 Sep
50.9K
369.4K
18.3K
51.4K
6.8K
8.9K
2.4K
12.9K
4294 Sep
4.6K9 Sep
149.5K
4.3K
5K10 Sep
Last 14 collected posts, oldest on the left. The scale is logarithmic: one post can outrun the rest a hundred times over.
Engagement rate per post
0.42%3 Sep
0.29%
0.89%
0.15%
1.24%
0.22%
0.68%
0.98%
0.20%
0.70%4 Sep
0.54%9 Sep
1.95%
0.30%
0.74%10 Sep
Reactions — likes, reposts, replies and quotes — divided by views. Median for 10K–100K accounts is 1.52%.
What the audience does
Likes63.5%17 562 in total
Reposts2.9%814 in total
Replies1.9%532 in total
Quotes0.5%143 in total
Bookmarks31.1%8 595 in total
Share of every reaction we collected for this account. Replies mean argument, reposts mean endorsement, bookmarks mean the post was worth keeping.
The follower curve appears once this account has two daily snapshots — we take one a day, and this one is on its first.
Comptes similaires
Necdet Saraç ✓
@necdetsarac
36.5K followers
· 10.3K posts
Yeardley Smith
@YeardleySmith
36.5K followers
· 5.6K posts
UNFPA East and Southern Africa
@UNFPA_ESARO
36.5K followers
· 24.7K posts
Gerald A. Griggs ✓
@AttorneyGriggs
36.5K followers
· 52.3K posts
Vrbo ✓
@vrbo
36.5K followers
· 23.1K posts
Wendolee Ayala ✓
@Wendolee
36.5K followers
· 20.9K posts
PUBLORD ✓
@publord
36.5K followers
· 41.5K posts
番茄大叔🇨🇳 ✓
@uncleTomat0
36.5K followers
· 23.6K posts