Abdallah Absi (@absilutely) header
Abdallah Absi (@absilutely) profile picture

build, reflect, repeat. building village (https://teapaw.org/t.co/7KFMkbwrFC) - the social graph api previously product @LinkedIn

San Francisco, CA
Joined December 2007
Photo by Abdallah Absi (@absilutely)Photo by Abdallah Absi (@absilutely)Photo by Abdallah Absi (@absilutely)Photo by Abdallah Absi (@absilutely)Photo by Abdallah Absi (@absilutely)Photo by Abdallah Absi (@absilutely)Photo by Abdallah Absi (@absilutely)Photo by Abdallah Absi (@absilutely)Photo by Abdallah Absi (@absilutely)
Abdallah Absi retweeted
Wherever you stand on this issue it seems to me that this guy is taking it seriously and educating himself to form a coherent position (in this case landing on a nuanced one that is within the realm of debate worthy positions among tech experts). our government would be well served to have all congresspeople do that instead of sleepwalking regulating or not-regulating the most important tech of our generation
Abdul El-Sayed is the Pause AI candidate:
32
42
7
1,190
88,524
Abdallah Absi retweeted
Introducing: OpenGranola 🔥 I built an open source meeting copilot for macOS. It transcribes both sides of your call on-device, searches your own notes in real time, and hands you talking points right when the conversation needs them. No audio leaves your Mac. Point it at a folder of markdown files, pick any LLM through OpenRouter (Claude, GPT-4o, Gemini, Llama), and it just works. It's invisible to screen share too — nobody knows you have it. The whole thing is open source. Link below
155
109
29
2,258
293,842
Abdallah Absi retweeted
While Cali man is is going poor setting up his 100 open claw swarm, Florida man is getting rich with a free ChatGPT sub
Florida man sold his house in just 5 days after letting ChatGPT handle the entire process instead of a real estate agent The AI handled pricing, marketing, showings, and even helped draft the contract
46
19
1
514
66,853
As a Christian who has lived in the Middle East for many centuries of accumulated history, I can say this clearly: the problem is not Islam. Islam, like Christianity, Judaism, or any other religion, emerges from a cultural and social structure. It is interpreted, practiced, and lived differently depending on the era, the level of education, the economic conditions, and the historical trajectory of each society. Religions are not static blocks; they evolve with societies. It is therefore intellectually dishonest to claim, in absolute terms, that Islam itself is the problem. First, Islam is not one single reality. There are multiple Islamic sects, diverse peoples, radically different cultures, and very different political experiences. Indonesian Islam is not Saudi Islam; Moroccan Islam is not Iranian Islam; Bosnian Islam is not Afghan Islam. To reduce all of this complexity to one label called Islam is not analysis، it is simplification bordering on propaganda. The real problem is far more complex and multidimensional. The first major rupture occurred during the Cold War, when the Soviet Union successfully penetrated Arab political structures under the banner of Palestine. This produced a form of leftist Arab nationalism, deeply hostile to Western civilization and ideologically mobilized against the United States. Over time, many of these movements failed politically and economically, and gradually recycled themselves into what can be described as Islamized leftism retaining the same anti-Western ideology, but now dressed in religious language. The second phenomenon is demographic and sociological. Due to declining birth rates, Western countries imported large numbers of Muslim migrant workers. Most came from rural areas or impoverished urban neighborhoods, often without education or cultural capital. These populations were neither fully anchored in their societies of origin nor successfully integrated into Western societies. This failure cannot be reduced to religious blame alone, it is the result of economic policy, urban segregation, and integration failure. This complexity was understood early on. Hassan II, the late King of Morocco, explicitly warned France against mass naturalization of Moroccan migrants and expressed serious reservations about the long-term consequences. His concern was not religious, it was sociological and structural. No serious thinker judges a religion based on isolated incidents, or based on periods of decline or prosperity. History shows clearly that wealthy and stable states integrate diversity far more easily than poor and fractured ones. Poverty radicalizes; prosperity pacifies. Finally, there is the most dangerous illusion of all: the idea of turning global tensions into religious wars. By what rational logic should humanity drag the planet into civilizational or religious conflict? This is not only immoral, it is strategically suicidal. Anyone advocating such ideas lacks historical depth and strategic vision. The conclusion is simple: Religious hatred is not analysis. Generalization is not intelligence. And blaming Islam as a whole is neither accurate nor useful. What we are facing is a crisis of failed states, ideological manipulation, economic exclusion, and strategic irresponsibility, not a war of religions. Those who truly care about peace must stop feeding fear, and start addressing reality.
202
194
35
997
145,101
Abdallah Absi retweeted
If you're inexperienced, don't try to pretend you're not. It will fool no one and make you look ridiculous. Instead just be openly curious. This will seem natural instead of awkward, and you'll learn a lot more.
209
1,163
111
12,668
524,921
Abdallah Absi retweeted
I love APIs, but as we move towards a more agent-based world, I think we'll see forward-thinking products and platforms offer an interface that's built for agents. Today, we have robots.txt and sitemap.xml support. They provide instructions to webcrawlers. Tomorrow, I can imagine us having an agents.json file on websites that provides a description of how to "talk" to the agent interface and what features are supported.  Kind of like a discoverable REST API. So, agents like OpenAI's Operator could then query that file first to see if there's an agentic interface, and if not, revert back to the less-optimal human-oriented one. What'll be super-interesting is how agents will self-identify who they are and which human they're acting on behalf of -- hopefully, in a verifiable way. Sort of like OAuth for agents. Fun times ahead!
215
185
64
2,040
271,854
Abdallah Absi retweeted
Whenever I go to a YC alumni meetup, one of the things that stands out is how many companies seem to be going nowhere for years, and then take off in years 4-6.
73
71
35
1,819
271,389
Abdallah Absi retweeted
Your product doesn’t need a lot of features but the thing that you’re testing needs to feel like a German car door—reaching a level of quality that doesn’t distort signal. Distortions can arise from the product being so half-baked that users don’t allow it to proceed into the consideration stage. Contrary to popular wisdom, sometimes a slick animation early on is the cheapest way to sneak into the consideration stage. This is especially the case if your greatest uncertainty lies at the top funnel.
42
109
20
2,076
237,122
Abdallah Absi retweeted
8 year olds can code like this. The future is so bright.
27
43
15
433
91,153
Abdallah Absi retweeted
“For you” killed the retweet.
127
147
41
2,991
514,347
Big congrats to our inaugural Flux cohort! This is the first time since the old Thiel Fellowship days that we have run a group program. Excited to work with these renegades!
9
7
2
100
6,701
China's foreign policy, so far, seems less destructive than the obsolete Anglo-American model. They try to get parties to get along, such as Iran & Saudi Arabia (or the various Palestinian factions); the Anglo-American model consists in backing one ag. the other. This is yuuge.
197
435
76
2,928
360,190
Abdallah Absi retweeted
Yesterday, Joe Biden included trespassing in his description of violent and disruptive protest, saying it was not okay. In 1960, four young black men refused to leave Woolworth's lunch counter in Greensboro, North Carolina, in violation of the law. They were trespassing on private property and being disruptive. MLK was arrested for the same thing in Atlanta. Those arrested were charged with trespassing. MLK was sentenced to 6 months hard labor. If Joe Biden were president then, he would have told the protesters to go home. JFK was a Senator running for president then- he actually made phone calls to get MLK out of his jail sentence. The point of protest is to be disruptive. Civil Disobedience literally means breaking the law. Kudos to the protesters of the 1960s and of today who understand what it takes for justice to prevail.
547
4,816
204
15,366
533,425
Abdallah Absi retweeted
Whatever your cause, if you damage art or block traffic to "draw attention" to it, you're actually hurting it, because you alienate everyone who'd been neutral up to this point.
292
167
182
2,837
1,158,089
Abdallah Absi retweeted
Most people who do evil believe they are doing good. So for everyone here who is absolutely certain that you are on the side of good should perhaps step back and consider the possibility that you are wrong. This is difficult, but necessary.
1,734
1,784
411
14,896
1,260,433
Abdallah Absi retweeted
I am so proud/excited/thrilled/delighted to share we've raised a $13.5M Series A at @withshepherd . 9 years ago, I moved to San Francisco with the goal to work in tech and one day start my own company. 3 years 2 months and 3 days ago, I quit Airbnb to start Shepherd with @justdlevine and Steve. At first, we didn't know "what" and "how" we were going to do... but we knew "why". We knew the insurance industry needed a new and fresh solution. That's what we're doing at Shepherd, and we're lucky to be joined by so many incredible individuals who shape this team. The Series A is a milestone in our journey; the hard work is only getting started. I am so excited for what's to come and all we'll accomplish! shepherdinsurance.com/blog/a…
89
9
4
516
71,177
Abdallah Absi retweeted
Here's a common startup situation. A team busts their ass for months building the first version of their product. It's almost done. Now a big question emerges -- how do you get the first people to use your product? Hmm... If you find yourself at this moment, then you are already in a bad place. 99% of startups are not differentiated on their underlying technology, and there is very little engineering risk involved. (I'm ignoring deep tech and foundational AI research companies, for the sake of this conversation). Because technology differentiation is no longer a real factor today start ups, it turns out that most products are succeeding or failing due to core product/market fit followed by the distribution strategy. There are over 9 million mobile apps. There are a billion websites. Figuring out distribution is key. This is why I think startups end up needing both: 1) an insight about customers that gives them product/market fit 2) an insight about distribution that creates traction People building products often have an easier time product/market fit because they are building for themselves, or a customer that they already know well. But the latter, about distribution, is often super difficult because once you onboard your friends and family, and look to expand the next set of hundreds of customers, you then dive into the world of growth marketing strategies and tactics which are its own very particular learned skill set. Sometimes when there's a new breakthrough technology, as with what is happening in AI, or the Apple vision Pro, or Web3, it's simply enough that the product has a "it works" feature. By simply being there on the scene when adoption of a new platform is happening, distribution happens automatically. I think that's why we see that so many new great startups are launched right at the beginning of the platform. But what happens when you are trying to launch the 9,000,001th mobile app? The first thing you do, naturally, is to try to read what's out there. The other counterintuitive thing, is that although most of the knowledge in writing out there pertains to channels like SEO or paid marketing or influencer campaigns, many of these tactics best fit already successful products that have money and aim to accelerate growth. Many of these tactics simply won't apply to you because they'll be too expensive, or they will use mature marketing channels that just won't be that effective. I often joke that by the time there's a case study about a new marketing tactic or channel, the advantage has already been arbitrage away, and probably no longer works. So what should you do instead? Ideally the product and the distribution hypotheses happen at the same time, and reinforce each other. The Dropbox founders describe to me at the inception of their product, that sharing folders was part of the vision and was built in quite quickly. And later years this drove a significant amount of growth. Uber has natural virality because you often ride in a car with other people, or you ride a car to see somebody, and naturally you'll mention the service. A product for creators, like Substack, will naturally encourage people on the platform to write and share content, attracting an audience who ultimately may also be writers themselves. Zoom, and other apps that help collaboration in the workplace, have natural features that cause you to bring in your coworkers as you use the product experience. These are all examples of the best form of distribution, which are baked in to the product idea itself, rather than bolted on at the end. Even once you have a basic theory for how your product will naturally distribute itself, you'll still need to identify the first generation of users to help iron out all the issues, and give you feedback on whether your hypotheses were correct. In my years of studying new product launches, I can confidently say that the early years are often very idiosyncratic, and constantly changing. The reason for this of course is that marketing channels change all the time, but subscale ones that help you get your first couple thousand users, change even more so. A few years ago you saw a trend were products would launch a huge conferences like SXSW. These days you see more effort on getting influencers involved early. Or "building in public" which makes yourself into an influencer. Several years ago many consumer products (like dating sites, new photo apps, etc) would launch on college campuses via the Greek system, because they were organized ways to reach thousands of undergraduate students. These days the organizations are often inundated with start up requests, and it's become less effective. As a result all of these initial channels change all the time, and it's up to the founders to figure out how to take advantage of what might work today. The problem with these initial channels is that they eventually tap out. Thus starts the journey of startups to grow and expand their portfolio of distribution channels, beginning with small and highly relevant ones, into the biggest channels. I sometimes imagine a X Y axis, where X is volume of the channel, and Y is responsiveness. Early channels are often very low volume. But you want that. The reason is that they are highly relevant and they are small enough that larger companies do not focus on them. As I mentioned influencers are often an example of this, but so are niche newsletters, or or event marketing. However if you find this channel to be successful, you'll also eventually one more scale. This involves you jumping onto the next set of channels, which will provide more volume but be much more competitive as a result. Often times this is a period where you have one channel that kind of works, and you're testing a few other channels simultaneously. Your efforts here should be experimental and iterative. You can often look at direct competitors as well as adjacent products and see what they're doing, to inspire you on the right channel. The natural cadence of products will indicate to you the channels that are most likely to work. If you have episodic usage, you'll probably need to do SEO/SEM, affiliate, or referral -- something that helps you target high intent users. If you're product is social or helps with workplace collaboration, then you might lean into referral programs and viral growth. Products in commerce naturally lead you towards paid ads, contact creators, etc. You can often learn a lot by talking to other people in your industry or an adjacent industries to see what works. This is where sometimes I'll see people working on episodic usage apps, like travel/health/etc asking the question, how do I make my product virally? I want free users! Of course the problem is, there's a natural fit between a product and it's distribution channels. Even though you might want free distribution, only very specific niches of networked products are able to grow freely. Generally everybody else must pay for their distribution, whether via referral or advertising. Eventually you want to move on the XY axis towards volume. There are only about a dozen large scale distribution channels that can propel a product to scale. Advertising is on that list, SEO too, and so is viral growth. But these larger channels, by their nature, are both highly scaled but also have low responsiveness. As a result, you end up competing with some of the most famous brands in the industry as a result. Who wants to buy ads against the same audiences as major credit card or airlines? They have insanely high payback periods, and huge marketing budget, and are not that cost sensitive. Ironically, this is where great products become to dominate. I started this discussion with the dual requirement of product/market fit, and distribution. But in the end, product/market fit actually dominates. The reason is the following -- the ability for a company to operate out in these most expensive and highly scaled channels comes from having a great product that generates a ton of word of mouth. More natural usage, the less marketing that has to be done. And the marketing costs that do exist end up being blended in with the large number of organic users. The journey of a new product is to move, from unscaled and relevant, to highly scaled. And at the end, great products win.
77
204
49
1,681
514,679
Why can't calendly links talk to each other and find time between people in 1-click. add conversational AI for confirmations/finding another suitable time
1
308
The latest PCPSR poll of Palestinian public opinion is out and I am going to share some of the highlights from this poll looking at Palestinian attitudes in this historic moment...
34
870
140
2,200
576,725
Abdallah Absi retweeted
Well, well, well. That felt coordinated. Turns out it was.
664
3,617
587
13,336
3,508,152