Christina, PhD

143 posts

Christina, PhD banner
Christina, PhD

Christina, PhD

@c_harker_

Head of Growth at an AI infrastructure startup. Love AI + GTM strategy. Obsessed with business economics, PMF, and how to stop startups failing. Yale PhD.

Katılım Temmuz 2024
374 Takip Edilen53 Takipçiler
Sabitlenmiş Tweet
Christina, PhD
Christina, PhD@c_harker_·
For a laugh I made this on a break at work. This is the power of niching at an early stage: your target zones should be outside the circle hit by standard industry solutions (even if you're building your positioning as a person and not as a product). Build for the edges and strengthen from there towards the centre. Ideas by @themgmtconsult and me. ;)
Christina, PhD tweet media
Maurizio@themgmtconsult

Being a category of one sounds daunting because we *think* it requires some singular world-class superpower. My idea is that it doesn't, and at the end of this post I'll give you an exercise you can run today. The key concept here is that you gotta find rareness at the intersection of 3-4 common skills. Practical example: > A great coder is common > A great coder who understands behavioral economics is rare > A great coder who understands behavioral economics and the specific regulatory hurdles of the healthcare industry is a category of one. When you combine these layers, you create a logic that your competitors cannot follow without dismantling their own business models. They might be better coders, or better economists, or better compliance officers, but they cannot be that specific combination. A lot of people try to build a "solution" and then go looking for a problem. I try to consistently do the opposite. Every complex problem has a specific jagged shape. Standard industry solutions are usually round: they are designed to fit as many holes as possible. They leave gaps. You need instead to look at the jagged edges of the client's specific pain and assemble your capabilities to fit that exact geometry. Admittedly, it doesn't always work... The "secret" to being the only is the willingness to be the wrong choice for almost everyone else. Incidentally, a book I gift to every new analyst I hire is "The Courage to be disliked": if you haven't read it, you should. Differentiation tries to please the whole market while being slightly different, but a category of one is designed to be perfect for a specific problem and useless for everything else. This isn't actually harder than traditional competition. In fact, I believe it's easier! Competing to be "better" is an endless (and frankly exhausting) treadmill. Instead, building a category of one is simply a matter of looking at your existing tools and *deciding* to arrange them in a way that solves a problem no one else is willing to touch. Do this exercise. Look at your own set of skills or your organization's assets: which 2 or 3 "standard" capabilities do you have that, if forced together, would make your competitors uncomfortable? Start from there. Let me know.

English
1
1
8
1.4K
Christina, PhD
Christina, PhD@c_harker_·
@QiaochuYuan pretty sure the person who most credibly claimed to protect us from demons was Mom 🤷‍♀️
English
0
0
4
492
Christina, PhD
Christina, PhD@c_harker_·
@mattworkman Oh wow, I thought this was a "what's the best career move" type of poll, not a rent vs buy one. Hmmm here's my thinking: I think LA is right if you're a normal human, SF if you have a huge deposit. Pasadena looks really nice tbh too. Both let you have jumbo loans so 🤷‍♀️
English
1
0
0
38
Matt Workman
Matt Workman@mattworkman·
I’ve looked at house/apt prices for both equally WILD, not that Boston is “cheap.” NYC is kind of an option but I imagine you are flown to CA several times a month and still pay Manhattan rent. Don’t want to live in BK. when my kids are in college I’ll have more freedom to move. If humans are still in control by that point
English
1
0
1
65
Matt Workman
Matt Workman@mattworkman·
what is more likely to succeed?
English
4
0
0
575
Christina, PhD
Christina, PhD@c_harker_·
@signulll we're all adults. we can admit we'd spend company money on SaaS tools that play la cucaracha.
English
0
0
0
41
Christina, PhD
Christina, PhD@c_harker_·
Wrappers basically can't have PMF because true PMF has some level of defensibility (moat). By definition, wrappers have zero. Most cynical view: you're getting something low value to them that seems high value to you so you do the PMM work for them and then they take your market. The only way around this would be LLM-agnostic feature that is embedded in secret sauce.
English
0
0
0
78
Ansh Nanda
Ansh Nanda@anshnanda·
You don’t have PMF if you are selling tokens for pennies on the dollar. If you have the discipline to properly price the product and take advantage of the tokens, then by all means. But most YC founders won’t realize this until it’s too late. Don’t be one of them.
English
5
1
28
3.6K
Ansh Nanda
Ansh Nanda@anshnanda·
DO NOT take the $2M OpenAI deal. You’re going to build a company that thinks it has PMF and just becomes a token reselling business. Here’s how 👇
Rohan Varma@TheRohanVarma

Yesterday @sama just offered to invest $2M of OAI credits into any current YC company as an uncapped SAFE. If I was in YC today, here is why I would take the deal: 1. You will spend that much $ on tokens building your product quicker than you think. The best AI-leveraged engineers are spending ~$10k+ per month on tokens. Thats $1.2M/year for a 10 person engineering team. The OAI deal means you don’t have to think twice about accelerating your engineering. 2. OAI tokens are worth 2X Anthropic tokens. Our frontier models are ~50% more token efficient than Ant’s. This means our $2M in GPT tokens is worth $4M in work accomplished with Opus. 3. You can use the tokens on the API. This means you can offer agentic products to customers without worrying about price or charging while you find PMF. Any product you offer that is useful with agents will use a non-trivial amount of tokens. Worrying about those costs while finding PMF doesn’t seem worth it. 4. The dilution will be minimal if you find PMF as an AI-native product and raise a Series A. A lot of series A’s I see these days are in the $100-$200M valuation range. $2M at that valuation is 1-2% dilution, which is very worth it if it allowed you to defer raising as much in your seed in order to pay for tokens. Would be curious how folks are thinking about it! Definitely an interesting offer to consider 👀

English
8
1
72
53.3K
peepeepoopoo
peepeepoopoo@DeepDishEnjoyer·
ai isn't going to take all our jobs. but it's going to take like, the bottom 20% of people's jobs. and that's a problem, since americans aren't as hot as spaniards or greeks
English
36
3
559
32.8K
Christina, PhD
Christina, PhD@c_harker_·
Popping up to say this applies to image and video gen models too, not just LLMs. The math will kill you money-wise at scale. It's absolutely brutal. If people can choose a SAGE (small and good enough) model for some tasks, they literally save tens or hundreds of thousands of dollars. 1k of Nano Banana 2 images costs you about $67. 1k of FLUX [klein] 4B costs $3. 1 million platform users needing an avatar thumbnail? Thats $67,000 vs $3,000.
English
0
0
5
852
Ideas Guy
Ideas Guy@nosilverv·
Must be crazy to be a woman — hairdresser fucks your shit up and then… what!? You can't even shave it bald!
English
5
0
16
1K
Christina, PhD
Christina, PhD@c_harker_·
It's so easy to fail correctly/discover and learn from non-starters: listen. Literally learn *how* to listen and then have conversations with your market. They'll tell you what needs to change, and what doesn't work about other people's products too. That gap is where you win. So many tools and methods for doing this, it stuns me that so few people seek them out. 🙉
English
1
0
0
304
tobi lutke
tobi lutke@tobi·
Failure* as an image problem. It’s the successful discovery of something that didn’t work. * non-catastrophic
English
15
23
340
26.6K
Christina, PhD
Christina, PhD@c_harker_·
The industrial ramp up enables cost efficiency, and that's foundational to the strategy. Ukraine exchanged infantry asymmetry for cost asymmetry in their favour.
Christina, PhD tweet media
English
0
0
0
29
Christina, PhD
Christina, PhD@c_harker_·
@Duderichy didn't the guy find his car or his bike later after it all settled down again? and get arrested for being beyond the fire line or some such? did I dream it? he basically did the ending of an action movie for the hell of it.
English
0
0
5
1.1K
Christina, PhD
Christina, PhD@c_harker_·
@parmita It never occurred to me that non-animal testing models might re/open paths for drugs that can't otherwise advance through the process. And I also support it because of animal welfare on top of the efficacy question fwiw. Good on you and what you guys are doing.
English
0
1
9
2.6K
Parmita Mishra
Parmita Mishra@parmita·
We are sending our public comments to the FDA today, for the MOST IMPORTANT FDA GUIDANCE in modern drug development. We have prepared the following comments as an overview. We will also be sharing our official comment documents on our website, link will be posted TODAY!
English
7
8
114
18.3K
ThePrimeagen
ThePrimeagen@ThePrimeagen·
I am not trying to be mean at all here, I am genuinely confused as to the purpose and I am afraid this will live rent free in my head for a while.
English
7
1
396
22.3K
ThePrimeagen
ThePrimeagen@ThePrimeagen·
I have reviewed the language and really tried to understand this, but I really do not understand this language's purpose other than engineers with too much free time, free tokens, and a marketing budget. I was very excited to read about a language is "agent's first." Its just zig with a touch of java and rust...
Chris Tate@ctatedev

Introducing Zero The programming language for agents. I wanted a systems language that was faster, smaller, and easier for agents to use and repair. Explicit capabilities. JSON diagnostics. Typed safe fixes. Made for agents on day zero.

English
188
35
2.3K
338.8K
Christina, PhD
Christina, PhD@c_harker_·
@lennysan @kalinowski007 One of the wildest things to me is the performance difference between chips optimized for particular types of models and ones not. Seen the difference with my own eyes for image and video gens. Crazy stuff.
English
0
0
1
217
Lenny Rachitsky
Lenny Rachitsky@lennysan·
My biggest takeaways from ex-OpenAI, Apple, Meta roboticist @kalinowski007: 1. The AI frontier is shifting from digital to physical because labs see the ceiling of keyboard-bound AI. “What you can do behind a keyboard with AI is going to saturate.” Which is why labs, big tech, and startups are increasingly investing in hardware, and why enrollment at universities is rising while CS enrollment is trending down. 2. More change is coming to warfare than to consumer electronics in the next two years. Drones, robotics, and the hardware supply chain all converge on the battlefield, and Caitlin argues we have to be able to control adversarial threats to our hardware layer, not just our chatbots. 3. The hardware industry faces a looming memory crisis that could derail the robotics revolution. Memory prices are spiking—potentially doubling or more—driven by AI data center demand. Companies building consumer robotics can’t compete on price with data centers. Caitlin is advising startups to pre-buy memory and stockpile components if they can afford it, because “we are in trouble as an industry.” 4. VR didn’t become a mainstream product, but it created the tech necessary for robotics (and war). SLAM (simultaneous localization and mapping), depth sensors, spatial computing, and understanding how humans perceive visual data in space is now powering robotics, autonomous vehicles, and drones. The technology needed to understand how a robot moves through space is essentially the same technology developed for VR headsets. 5. Humanoid robots are overhyped. While humanoids are interesting for certain long-tail tasks, most manufacturing and real-world applications need dedicated robots designed for specific jobs. A robot screwing keyboards into laptop cases doesn’t need to be humanoid—it needs to be optimized for that exact task. The future will have robots for construction, electrical work, logistics, and low-volume assembly, and most won’t look humanoid. 6. Supply chain independence is a national security imperative. Over the past 25 years, essentially every layer of the hardware supply chain—from raw magnets to actuators to final assembly—has been outsourced to China, Japan, and Korea. The same actuator technology that makes a drone rotor spin also makes a robot arm move. Without an independent supply chain, the U.S. is vulnerable. As Caitlin warns, “We need to re-industrialize this country significantly in order to be safe in a military sense.” 7. The hardest part of building safe robots is the decisions you don’t think about. If a robot arm is heavy and hard, the impact force when it hits you is dangerous. But there’s also the social aspect—robots need to show intent before moving (looking before turning), acknowledge when humans enter a room, and transmit non-threatening body language. As Caitlin learned from researcher Leila Takayama, “If a robot just suddenly turns and does all this stuff, it scares you. But if a robot looks before it turns and then goes, it’s much less alarming.” 8. Software builders don't understand how fundamentally different building hardware is. Software can compile code hourly, but in hardware you may get only a handful of chances to “compile” before mass production, with each major build taking three to five months. Once you ship, you’re done—there are no over-the-air updates for physical components. Software intuition doesn’t transfer to hardware. 9. In hardware, you never have enough time—so do everything you know you need to do right now. Caitlin learned from Apple executives like Shelly Goldberg and Kate Bergeron that you can’t wait around. Even if you technically have more time, use it, because “in two days there’s going to be a surprise coming around the corner that you need that time to fix.” This ruthless efficiency of clearing known tasks immediately creates a buffer for inevitable surprises. 10. AI hasn’t yet transformed hardware engineering. AI can’t do real CAD (computer-aided design) yet, and AI models don’t understand friction, weight, contact pressure, or surface texture. It can do surfaces and point clouds, but not the dense, equation-based solid entities that hardware engineers need. But when it arrives, it will be transformative. 11. CAD files are some of the most valuable IP any company has. Samsung, Apple, and other manufacturers will never give their 3D CAD to AI model makers. This creates a data scarcity problem for training hardware AI. The solution might start with hobbyists who don’t care about IP protection and just want to build things faster, then eventually move to on-premise AI systems that companies can train on their own data without sharing it externally. 12. The best hardware teams combine three types of people. You need generalists who can apply lessons from other fields to new problems. You need some specialists who have built similar things before and others who have scaled products to high volume. And critically, you need 20-year-olds who are truly AI-native—they approach problem-solving completely differently because they use AI from the ground up for everything. As Caitlin notes, “It’s very hard to find someone who’s in their 30s who can be truly fully AI-native.”
Lenny Rachitsky@lennysan

Caitlin Kalinowski (@kalinowski007) helped engineer the original unibody MacBook Pro and was technical lead on the MacBook Air and Mac Pro at @Apple, was @Meta's first consumer electronics hire and went on to lead their AR glasses and VR hardware teams, and most recently was at @OpenAI helping build their robotics and hardware teams from scratch. In our in-depth conversation, we discuss: 🔸 Why the AI frontier is shifting from digital to physical 🔸 How the technologies built for VR became the foundation of modern warfare 🔸 Why humanoid robots are still just prototypes, and what’s most gating mass deployment 🔸 The coming memory price shock and why she’s telling startups to pre-buy now 🔸 Lessons from Steve Jobs, Mark Zuckerberg, and Sam Altman 🔸 Why she walked away from OpenAI after the DoD deal Listen now youtube.com/watch?v=G5WTgB…

English
8
28
190
34.7K
Christina, PhD
Christina, PhD@c_harker_·
@lulumeservey This is the opposite of empowering for these kids smdh. They should be hearing about how they can be the ones building world changing technology with AI.
English
0
0
2
645
Lulu Cheng Meservey
Lulu Cheng Meservey@lulumeservey·
“The question is not whether AI will shape the world; it will. The question is whether YOU will help…” Tonedeaf disaster Obvious move would’ve been to focus on the graduates as the protagonists, instead of framing them as accessories to AI’s world takeover “The question is not whether you will shape the world; you will. The question is how…” simple switch Then you can go on to talk about adapting to change, trying new things, using the power of technology newly available to them. And make it about empowering new graduates Instead it ended up preachy, condescending, and vaguely menacing all at once
Alex Kantrowitz@Kantrowitz

This is incredible. Artificial intelligence getting booed out of the stadium in any commencement speech it’s mentioned. Maybe telling college students AI was taking their jobs wasn’t the best strategy. Must watch —>

English
63
24
481
73.8K
INFOSEC F0X 🔥
INFOSEC F0X 🔥@infosec_fox·
@c_harker_ Didn’t know it was called Sefa genesis in the U.S. I had the SNES too and had that bazooka
English
2
0
1
56
INFOSEC F0X 🔥
INFOSEC F0X 🔥@infosec_fox·
AGE CHECK: Who had a SEGA Megadrive 1? 👀
English
18
0
20
1.1K
Brooke LeBlanc
Brooke LeBlanc@brookeleblanc·
What a beautiful day to lay in the park and not be hungover by a depressant, neurotoxin and class 1 carcinogen.
Brooke LeBlanc tweet media
English
11
1
158
7K