Sabitlenmiş Tweet
i'm starting my newsletter on @beehiiv called startup spells 🪄
join in to get the first issue.
the newsletter will cover startups & marketing tips in a short 5-min daily read with a badass curations.
startupspells.com
English
Akshay from Startup Spells 🪄
30.2K posts
Akshay from Startup Spells 🪄
@StartupSpells
I cover 0.1% Marketing Strategies of Startups from The Past, The Present & The Future. Get them daily in your inbox at https://t.co/kk9u7nveNp
📩 Join my free newsletter Katılım Aralık 2011
1.8K Takip Edilen2.2K Takipçiler
@Linahuaa I love ur takes. Very unique & hilarious.
English
All Asian countries have low birth rates due to overpopulation, intense competition.
But Korea is extra special because of its traditional Christian culture AND high-level Western feminism at the SAME time, driving the largest ideological divide between men and women on the planet.
And on top that, they have the worst role models- they are a celebrity factory, but like none of the actors and singers even get married before 40, let alone get children.
And on top of THAT they also have the most extreme looksmaxxing of any country on the planet by far.
BMI > 18 means you're a fat pig.
Women are super scared of getting children (=fat and ugly)
Crémieux@cremieuxrecueil
The UN projection for South Korean fertility is wildly, insanely, gobsmackingly optimistic. Come on, man.
English
@GremlinIndustry @r00k They do.
Insta/Tiktok//makeup industry or even society uses manipulation techniques on women (young or middle aged) that u need plastic surgery (even beautiful miss worlds)... just different techniques tho.
Nature is balanced if u look closely.
English
@StartupSpells @r00k Then they should be telling different lies and using different manipulation techniques on each sex
English
The fact that dating apps present the same interface to both men and women seems completely wrong to me. These are not the same kind of user! Their UIs should be built around totally different assumptions. They practically need to be separate apps!
English
@daymakerguy Genius idea lol!
There's many verticals for this.
English
Last month we invoiced $10k in cakes
This month we’re invoicing almost $100k beacuse coldcaking blew up.
cold emailing and cold calling is officially cooked and baked.🎂👨🍳
English
🚨 BREAKING: OPENAI PRE-IPO FILING LEAKED
OpenAI just told investors Microsoft is a risk to their business and they are diversifying away from them
>be microsoft
>give openai your money
>give them most of your compute
>become their primary distribution partner
>their largest shareholder, 27% equity ($135 billion)
>then officially list them as a competitor to microsoft
Satya Nadella to microsoft board:
"Don't worry if OAI disappeared tomorrow, we have everything we need. All the IP rights, the capability, the data, the compute... EVERYTHING. We are below them, above them, and around them."
OpenAI pre-IPO filing to investors:
“If Microsoft modifies or terminates its commercial partnership with us, or if we are unable to successfully diversify our business partners, our business, prospects, operating results and financial condition could be adversely affected”
the funniest part of all of this is 45% of microsoft cloud contracted backlog is driven by openai api… but openai just signed a $50B deal with amazon that violates azure exclusivity lmao
Microsoft now threatening legal action against OpenAI
most expensive breakup
ITS HAPPENING
Elon Musk@elonmusk
@satyanadella OpenAI is going to eat Microsoft alive
English
@emollick @MParakhin I think they started hiding traces for Chinese models otherwise they'd reverse-engineer it by asking questions.
Otherwise, Google used to show it before.
English
@MParakhin Its a great model with a bad harness that makes it hard to use for many tasks.
Ethan Mollick@emollick
The new Deep Think is very, very smart, but is constrained by the Gemini interface compared to GPT-5.2 Pro. If I can't see actual work in the thinking trace, or get downloadable files as proof of work, or see evidence of the code & statistics it applied, checking results is hard
English
Incorrect: Gemini DeepThink is in the same class. It is more limited by the thinking budget, but I have to use both to get best results. Yesterday on one of the problems DT produced a much better solution - GPT-5.4 Pro agreed! “Your friend’s version is better” :-)
chatgpt.com/s/t_69c193e012…
Ethan Mollick@emollick
GPT-5.4 Pro continues to be the only model of its class. For anything really hard & complex, I throw it into the maw with every bit of context I can think of. More often than not, something very useful comes out. I can't get the same results from Codex or Code or anything else.
English
@PromptSlinger @paulg @nikitabier If you can prompt well, no one will catch you either.
Most AI slops are dumb because no one knows to prompt well obviously but its doable now.
English
@paulg @nikitabier wrote a regex to filter the 'great post! here's my take' pattern. caught 400 accounts in a week. then realized 12 of them were real people who just talk like that
English
Now before I reply to someone I have to check first that I'm not being baited by an account that spams everyone with AI-generated replies.
Can you write some software to do this check for me @nikitabier?
English
@youyuxi AI Slop has ruined the world. (jk jk lawl)
English
@Yuchenj_UW > Curious
Lol, why curious? They will.
English
OpenAI acquired Astral, the team behind uv, ruff, and ty.
Fun fact: Claude is the #6 contributor to uv.
Curious if Anthropic will ban them from using Claude since the team is joining OpenAI.
Congrats to the Astral team who built incredible Python tools!
English
English
English
We migrated @v0’s codebase from ESLint + Prettier to Oxlint + Oxfmt by @oxcproject, CI is now 3x faster. Love the OSS ecosystem.
English
@shadcn > Work that, for now, AI can't do.
How come? I did tell it to follow WCAG guidelines + contrast ratios & sometimes use Puppeteer & it worked but for different problem.
English
To get translucent menus right, we had to manually test components over different gradients & colors, checking visuals & accessibility. Now multiply that by 5 component types across 5 styles in both light and dark modes. It's a lot of work.
Work that, for now, AI can't do.
English
@noampomsky its called making decisions fast.
good workout session or good hobby = healthy mind.
healthy mind = u can work faster.
but nothing matters if u cant make decision fast. speed to decision needs to be instant.
difference between rich & poor is speed to decision.
English
as a rule the psychologically healthier someone is, the less time it takes them to do something they were eventually going to do anyway. Anything from sending a text to quitting a job. I’m not sure what to call it—processing time? lack of blocks? low avoidance?
English
@grayap9 @protosphinx Lol, he's ok. Most Indians anyways equate white = good-looking.
English
True, but if an Indian founder did this on the racism app the comment section would turn into the ten plagues of Egypt
Jeffrey Wang@jeffzwang
Roy Lee has done more for Asian American representation than any Olympian or Hollywood star
English
@thsottiaux Ask it not to read sensitive files
English
I've got openclaw managing my self hosted media center its so cool i can just do shit like this
English
They got a lot of hate.
But they were just early.
Higgsfield is the same premise- and is at 100 million arr
Ai is still coming for all content.
It will be 80% of an ad account within 12 months.
@levelsio@levelsio
Icon, the AI Admaker, just went bankrupt They paid $12M for the domain Icon.com and now it's dead
English
@burakeregar Woah, looks good. What happened to soreel?
English
my little app made it to the app store top 100 charts 🤩🤩
shifting my focus to marketing started paying off already 🎉
Burak Eregar@burakeregar
i literally stopped building after i sent the quoted tweet i am just taking some days off and focusing on marketing it started paying off. it's the first time i am seeing this much traffic! my UA cost is too high though, i need to see how this will affect the keyword rankings
English
@johnrushx interesting comment.
are blind humans as intelligent as non-blind? do u know anyone blind who's wicked smart? also, braille teaches them, no?
English
I dont share Yann LeCun's point of view
If it's true, how can he explain the intelligence in blind humans?
They have never seen even a pixel in their life, but they are often as smart as a non-blind person.
He makes a typical mistake scientists/devs make by being ego-stubborn
Rohan Paul@rohanpaul_ai
Yann LeCun (@ylecun ) explains why LLMs are so limited in terms of real-world intelligence. Says the biggest LLM is trained on about 30 trillion words, which is roughly 10 to the power 14 bytes of text. That sounds huge, but a 4 year old who has been awake about 16,000 hours has also taken in about 10 to the power 14 bytes through the eyes alone. So a small child has already seen as much raw data as the largest LLM has read. But the child’s data is visual, continuous, noisy, and tied to actions: gravity, objects falling, hands grabbing, people moving, cause and effect. From this, the child builds an internal “world model” and intuitive physics, and can learn new tasks like loading a dishwasher from a handful of demonstrations. LLMs only see disconnected text and are trained just to predict the next token. So they get very good at symbol patterns, exams, and code, but they lack grounded physical understanding, real common sense, and efficient learning from a few messy real-world experiences. --- From 'Pioneer Works' YT channel (link in comment)
English