Domo

619 posts

Domo

Domo

@embossedly

Присоединился Temmuz 2025
21 Подписки8 Подписчики
Domo
Domo@embossedly·
@rand_longevity my belief is that things will remain slow until agi and then explode the last mile problem & human in the loop problem must be fully SOLVED for stuff to accelerate beyond current lvls that's why current systems aren't causing the world to change much, they're just narrow
English
0
0
0
141
Rand
Rand@rand_longevity·
i think we have 18 months left of the "old world" before the end of human of labour you can work if you want to but you will not have to and your job probably wont exist anyways
English
65
17
174
6.8K
The Kobeissi Letter
The Kobeissi Letter@KobeissiLetter·
BREAKING: Bitcoin falls below $64,000. Bitcoin has now lost -$400 billion in market cap since May 11th.
The Kobeissi Letter tweet media
English
330
381
3.1K
251.1K
Domo
Domo@embossedly·
@kiwitalkz 1B on marketing seems unrealistic. even Endgame only did 150-200M on marketing, i doubt TTWO needs to spend anywhere near that much given the game hypes itself even among casual audiences. it'll probably be a 100-200M campaign
English
1
0
1
766
Reece “Kiwi Talkz” Reilly
Reece “Kiwi Talkz” Reilly@kiwitalkz·
I want to explain something to Nintendo fans that seem to live in some delusional mindset where they think Nintendo won't be impacted at all by GTA 6. GTA 6 will be the biggest launch in entertainment history, period. Rockstar will be spending close to a billion dollars on marketing, you are going to see the game everywhere before release and even after release. It's going to be all over Tik Tok, streamers, youtubers and website publications will be covering it non stop, word of mouth will be all about GTA 6. Now obviously Nintendo diehards will buy Nintendo games regardless of this but Nintendo isn't just thinking about them, they have to market to people outside twitter, reddit, gaming forums bubbles etc. Now you can't do that when a single game is dominating the conversation everywhere, this is the part that Nintendo fans fail to realise, this is why publishers would rather risk cannibalizing each other in September than going head to head with GTA 6. Xmas is the most important time of the year for Nintendo and they will have to spend three times as much in marketing to even be close to piercing through the noise of GTA 6 especially in the U.S. their largest market. Nintendo will have to release something for the Xmas period no doubt and I am sure whatever it is it won't flop but they will definitely be impacted in some way by GTA 6 despite it not even being on the platform because at the end of the day attention is the fundamental currency of the entertainment industry. Something to think about.
Reece “Kiwi Talkz” Reilly tweet mediaReece “Kiwi Talkz” Reilly tweet mediaReece “Kiwi Talkz” Reilly tweet mediaReece “Kiwi Talkz” Reilly tweet media
English
174
38
486
125.5K
Domo
Domo@embossedly·
@deredleritt3r @danshipper what about xhigh? according to deep-swe and other benchmarks, the max effort has ~no gain over xhigh, so it seems like a waste
English
0
0
0
4
prinz
prinz@deredleritt3r·
@danshipper The Max reasoning effort is a game-changer and actually makes Claude somewhat useful for difficult reasoning tasks. Results from my private benchmark (prinzbench) for Opus 4.8 are coming soon.
English
2
0
29
2.4K
Dan Shipper 📧
Dan Shipper 📧@danshipper·
Almost a week later! What are your thoughts on Opus 4.8? We were extremely bullish on it in testing—it seems the response was more tepid once y'all got your hands on it. If you disagreed with our take I'm curious why so we can tune our evaluations! One theory I have is that by nature it pushes on your frame a little more, and the results are high-variance—sometimes it does something amazing, and sometimes it disagrees in a way that is obviously wrong. But curious how you're feeling and what you're reaching for after a few days of testing
Dan Shipper 📧@danshipper

BREAKING: Anthropic just dropped Opus 4.8—and it is a MONSTER We've been testing for about a week @every and our verdict is they could've just called it Opus 5, it's that good. Here's our vibe check: - Beats GPT-5.5 on Senior Engineer bench. On our toughest benchmark Opus 4.8 scores a 63—a hair higher than GPT-5.5's score of 62, and a full 30 points higher than Opus 4.7. It tackled a ground-up rewrite of a production codebase, and actually built something that works. HOWEVER: Coding performance varied a lot at different reasoning levels. We recommend using it on xhigh for best results. - Incredibly good writer. Opus 4.8 scored a 79.6 on our writing benchmark—measuring models on real-world writing tasks we do all of the time like essay writing, promo email writing, and more. It beats GPT-5.5 by 6 points. It produces well-written prose with fewer "AI-isms". It's also very good at writing in your voice given the right context. HOWEVER: Writing performance also varied with reasoning levels. Medium reasoning had higher incidence of AI-isms—we found best results with high. - Beast at knowledge work. Opus 4.8 is very good at general knowledge work tasks like report creation, research and more. It produced the best PowerPoint one-shot we've ever seen on our deck generation benchmark. - Emotionally intelligent, willing to question the frame. I've also found it to be quite good at talking through psychological or interpersonal issues. It has a high EQ, and it's also good at not glazing and helping to expand your perspective. Its thought process feels extremely rich and dynamic. THE BAD: These days a model is only as good as its harness, and Codex is still a far superior harness to the Claude Desktop app. This has kept me using Codex + GPT-5.5 as my daily driver, but I am flipping back and forth a lot more between Codex and Claude. Anthropic is back baby! Read the rest on @every: every.to/vibe-check/opu…

English
86
3
114
55.9K
Domo
Domo@embossedly·
@randomdude22401 @synthwavedd @kimmonismus the only model they've released on friday was o3-mini. dont get ur hopes up for now im believing polymarket which indicates next week prolly june 11th
English
1
0
0
100
welt
welt@randomdude22401·
@synthwavedd @kimmonismus friday - final answer I think it makes sense to do new models on fridays in general.. people have more free time to use it and post on social
English
2
0
1
846
Domo
Domo@embossedly·
@Ananthr27104587 fella we can literally see polymarket indicating june 9th or june 11th which has yet to be wrong on a gpt 5 update
English
0
0
1
8
Ananth
Ananth@Ananthr27104587·
Don't expect GPT-5.6 this Thursday
English
27
2
101
8.6K
Domo
Domo@embossedly·
@argofowl i remember people said use up all your usage the day before 5.5 came out because "they'll reset it" ... they never did don't make people panicwaste their usage ffs
English
0
0
3
42
🥔🥔🥔
🥔🥔🥔@argofowl·
make sure to /fast on xhigh today in codex a reset is very very likely with all these issues - i don't see how a reset isn't coming if it won't we will of course riot you know the drill
corey.ching@coreyching

@thsottiaux @romainhuet Thanks for flagging the image gen 2 error - flagging this internally. Please hold everyone.

English
15
3
97
12.7K
Domo
Domo@embossedly·
@JinjingLiang > The Codex app actually looks good it's looked good for ages because humans are steering it... GPT-5.x has been "good" at UI for ages if you have someone who knows what they're doing steering it only useless vibe coders who can't design for shit think GPT is bad
English
1
0
0
466
jinjingliang
jinjingliang@JinjingLiang·
GPT-5.6 is going to be very good at UI. My evidence: 1. The Codex app actually looks good. Much better than anything GPT-5.5 has made for us. They must be using GPT-5.6 internally. 2. OpenAI just shipped “Sites.” You don’t ship a feature for publishing AI-generated UIs unless you’re pretty confident the model can make good UIs. 3. GPT-5.5 is already strong at almost everything except UI. UI is the last obvious gap.
jinjingliang@JinjingLiang

Means GPT-5.6 is dropping any day now

English
70
31
876
120.4K
Domo
Domo@embossedly·
@joshbuildings @hsienshoryu @deredleritt3r already all over X, people are saying "ChatGPT reached 1B MAU!!!" including literally prinz himself you're saying: web + mobile: 900M *W*AU still mobile only: 1B *M*AU, new milestone just very confusing and no one layperson will know the difference here
English
0
0
1
24
Domo
Domo@embossedly·
@joshbuildings @hsienshoryu @deredleritt3r ugh idk the reporting is very ambiguous if you say "app" you need to say "mobile" with it to be specific that it's limited to iOS + Android, because "web apps" are a thing and what ChatGPT is
English
2
0
0
17
prinz
prinz@deredleritt3r·
Reuters: ChatGPT reaches 1 billion monthly active users. Congrats, OpenAI!
prinz tweet media
English
9
24
346
12.6K
Domo
Domo@embossedly·
@hsienshoryu @deredleritt3r nvm this is just according to some slop analyst firm called Sensor Tower, not from oai themselves i guarantee they've been at 1B MAU for months, since last year around Q4 at latest
English
0
0
0
14
Domo
Domo@embossedly·
@hsienshoryu @deredleritt3r no... it's definitely web + mobile combined, and their only other product is codex w 5m users and many of those probably already use chatgpt you would think 900M weekly uniques was already well over 1B monthly uniques, but... it wasn't
English
2
0
0
49
Domo
Domo@embossedly·
@OrenMe literally no one is using this in the big june 26 when codex and claude offer actual practical limits github copilot should pack it up and move on, it will never be relevant now
English
0
0
1
188
Domo
Domo@embossedly·
@GWHayduke97 but it got 3.5K likes and he's still at his shtick to a degree (pivoting to financials rather than capabilities) so we can continue to trust his insights <3
English
2
0
8
1.3K
Domo
Domo@embossedly·
@haider1 early April when Anthropic announced Mythos was the last we felt a step change (5.5 wasn't over 5.4 imo) people become accustomed to the new norm extremely fast now, so 2 months is ages and it doesn't help that all the new Opus releases are mid af, so progress has felt slower
English
0
0
0
134
Haider.
Haider.@haider1·
zoom out, even just a little look at mythos, claude cowork, codex, openai models solve maths Erdos problems, and everything around it i think people still underestimate the pace because it doesn't look as obvious as the jump from gpt-3 to gpt-4 AI progress is still a jagged line, not a smooth one yet but once AI can self-improve, even with physical limits, that line probably starts getting smoother and faster
English
8
5
53
3.9K
Domo
Domo@embossedly·
@edzitron these are not the "actual costs" the labs are forcing microslop to pay full api prices which is not the real compute costs (which are 5-10x cheaper) so using $2000 in api credits is more like $200-$400 in real costs which is why first party subs work unit wise
English
3
0
3
1.5K
Ed Zitron
Ed Zitron@edzitron·
If you want to see what happens when people have to pay the actual costs of AI, the day is finally here. It's obvious that every customer sees the deep, meaningful value and isn't angry at all
Ed Zitron tweet mediaEd Zitron tweet mediaEd Zitron tweet mediaEd Zitron tweet media
English
14
74
1.1K
70.9K
Ed Zitron
Ed Zitron@edzitron·
Day one of GitHub Copilot token-based billing and the customers LOVE IT! They're all celebrating the power and value of generative AI!
Ed Zitron tweet mediaEd Zitron tweet mediaEd Zitron tweet mediaEd Zitron tweet media
English
148
569
5.7K
1.2M
Domo
Domo@embossedly·
@alexecutedev @jasperdevs true but at worst its an 8-12 hour differential unless you dont use it after it resets. so close enough; at least a million devs are probably within 10 min of each other
English
1
0
0
11
Anderson
Anderson@alexecutedev·
@embossedly @jasperdevs On codex no one has the same end of the week 'cause the week start only at the first consumption
English
1
0
0
21
Domo
Domo@embossedly·
@parrapower2022 houses should not be an investment in the big 2026 just invest in ETFs
English
0
0
0
82
ParraPower
ParraPower@parrapower2022·
$11,000 growth over 11 years. The Australian Labor Government wants young Australians to build wealth patiently. Aussie Aussie Aussie Oi Oi Oi...
ParraPower tweet mediaParraPower tweet media
English
27
5
119
23.3K

Открыть

@rand_longevity @KobeissiLetter @kiwitalkz @deredleritt3r @danshipper @randomdude22401 @synthwavedd @kimmonismus