Kalam

8.2K posts

Kalam banner
Kalam

Kalam

@kalamishere

Compute for all • Sound Obsessed, AI XR, ex venture -Native Instruments

Katılım Ekim 2008
2.8K Takip Edilen1.7K Takipçiler
Kalam retweetledi
Robert Scoble
Robert Scoble@Scobleizer·
Reshares (formerly known as retweets) are dead. An end of an era in social media.
English
67
27
245
39K
Kalam retweetledi
arian ghashghai
arian ghashghai@arian_ghashghai·
i maintain the view that space will be the next big VC bubble after AI: > AI (like all bubbles) will eventually come to back down to earth > Losses in AI will be explained by it being infeasible to scale enough compute terrestrially quickly enough > Solution: put our AI data infrastructure in space (i.e. not on land) > Putting and maintaining exorbitant compute resources requires all kind of new investment in the space economy > space becomes the new hot, money pit (to fulfill the promise of AI) (it’s worth noting that VC has grown to big, and structurally requires bubbles to sustain its AUM)
Philip Johnston@PhilipJohnston

Chamath is right. “It’s the beginning, of the beginning, of the beginning” of the space industry. @chamath

English
23
6
89
23.5K
Kalam retweetledi
Andreas Klinger 🦾
Andreas Klinger 🦾@andreasklinger·
Imho twitter shut turn off payments for political posts. I dont mind people talking politics. I mind incentivizing people with money to say political flamespeak that gets the most engagement. It's like paying people to start barfights in a community pub.
English
4
6
86
6.2K
Kalam retweetledi
pollinations.ai
pollinations.ai@pollinations_ai·
merged 9 PRs today. shipped a CatGPT discord bot, fixed the sirius elevator's responsive backgrounds (now itch.io compatible), and added 6 new community projects to the greenhouse, including an AI RPG engine. the ecosystem grows.
pollinations.ai tweet media
English
0
1
2
89
Kalam retweetledi
Nathan Benaich
Nathan Benaich@nathanbenaich·
BIG DAY - GOOGLE MAPS LETS YOU SEARCH YOUR PINS!!!!
Nathan Benaich tweet media
English
9
4
83
8.4K
Kalam
Kalam@kalamishere·
Early Sora gem. Just before the rails and limits
English
0
0
0
8
Kalam retweetledi
pollinations.ai
pollinations.ai@pollinations_ai·
24 prs merged. wired up amazon nova, qwen, grok, and wan2.2 video. added device auth (rfc 8628) for headless cli logins. also moved flower & nectar tier refills to hourly, because waiting a full day to generate a pixel bee is objectively too long.
pollinations.ai tweet media
English
0
1
2
152
Kalam retweetledi
𝚔𝚊𝚒 🤖+🧠
𝚔𝚊𝚒 🤖+🧠@kaigani·
Can Claude Code create an entire animated short episode? My latest CODEYWOOD demo.
English
0
3
3
459
Kalam retweetledi
andrew chen
andrew chen@andrewchen·
There's a vast marketing industrial complex of agencies/consultants/advisors/whatever that promotes tech startups spending billions of dollars of unaccountable marketing budget. They're triggered by my anti-paid stance but here's the reality: - paid marketing is much, much worse than organic on every metric (conversion, ROI, etc) - startups work on a fast time scale and can't manage LTV/CAC correctly beyond a months timeframe - risk is asymmetric. a few bad cohorts can kill you (and btw, this has definitely happened) - the age of easy/cheap ad inventory is over. Pricing is controlled by an oligopoly, it's all being algorithmically bid up, and ROI sucks at scale - paid UA has S-curves. Early spend looks good, but plateaus and it's easy to get addicted - if your product is growing organically already, you might just be cannibalizing and pulling forward demand you'd already get anyway - high reliance on paid indicates weakness in the core product and value prop - you can't build a 100m+ DAU product with the majority coming from paid UA (it's just obv math) - going majority paid UA makes it 10x harder to raise VC capital down the line. For all the reasons on this list the main benefit of paid is simple: your agency/consultant/whatever spends money, some numbers go up, and you feel like you're doing something. It's simple to understand, you can apply it to every type of product, and every big co does it right? Billions of dollars swap hands just based on this dynamic. But for startups I argue it's the growth lever of last resort, since it's the most commoditized form of distribution -- you should try to exhaust your other ideas, invest deeper in your product, and grow based on whats unique in the ways that only your startup can grow. That way your channels are as defensible as possible, built around your killer value prop After all one day, you hit your CAC ceiling, your channel saturates, or worse, your competitors just do the same, copying your distribution strategy, dragging the whole industry into a prisoner's dilemma. When that happens, it's hard to incubate a bunch of new 0-1 channels to save your forecasts. The temptation is just to stretch payback periods, buy more, and ride it out. That's a dark path...
English
94
36
469
158.9K
Kalam retweetledi
Kalam retweetledi
𝚔𝚊𝚒 🤖+🧠
𝚔𝚊𝚒 🤖+🧠@kaigani·
I had Codeywood make an explainer video for Codeywood. No direction from me, all local models.
English
0
1
2
405
Kalam retweetledi
Naval
Naval@naval·
Software will proliferate just as videos, music, writing did. The market structure will shift from a “fat middle” to mega-aggregators and a long tail. It’ll be a slower process due to network effects, but many traditional vendor lock-ins will get eaten by AI.
English
656
731
9.9K
1.1M
Kalam retweetledi
Christian Steinmetz
Christian Steinmetz@csteinmetz1·
Thinking AI music begins and ends with typing a prompt and hitting generate is kind of like saying playing guitar is only strumming a few open chords. Sure, you can do that, but you can go much deeper if you want.
English
1
2
13
851
Kalam retweetledi
Yishan
Yishan@yishan·
There's a fair number of people who might have started college majoring in computer science but then switched to another field after a couple years, or people who graduated with a computer science degree (or minor) but moved quickly into another field. These people are now ULTRA POWERFUL in the era of vibe coding. If you are someone who has a background in programming but moved on to something else, vibe coding will tie it all together and give you amazing leverage.
English
17
17
339
43K
Kalam retweetledi
Bojan Tunguz
Bojan Tunguz@tunguz·
We just entered the final season. Of everything.
English
36
27
383
20.5K
Kalam retweetledi
James Jackson
James Jackson@derJamesJackson·
We raised the full amount within 2 hours, but I'm keeping it open to fund more independent journalism watching the hidden institutions of the state
English
2
8
117
2.2K
Kalam
Kalam@kalamishere·
@steipete native speakers struggle too - you'll be fine
English
0
0
0
27
Peter Steinberger 🦞
Peter Steinberger 🦞@steipete·
omw to London for an english language test Hope I’ll make it!! 🇬🇧
English
353
41
3.7K
335.5K