I don’t know if font foundries are 100% cooked, but watching Opus 5.5 methodically create whatever font you want…
While browsing a foundry where they try & price discriminate based on *industry* (as well as company size, font uses, usual nonsense)…
Certainly makes you wonder!
I would suggest everyone to try and and make a new font using Claude
it works so fucking good and u don't need to pay for commercial fonts anymore
maybe i should drop a tutorial/skill on this
1
4
502
literally luke retweeted
Replying to @arlanmarat
Yesss. Tried this a couple of weeks ago.
Worked with a sans serif font I was going for.
Didn’t work at all for a complex decorative display serif I wanted… at first.
Then gave it a similar open-source font as a starting point - got MUCH closer.
Ultimately made me appreciate the inspo font though ^_^
literally luke retweeted
welcome to life kids, here’s your screen. you absolutely have to learn how to use this thing because your entire professional and social life will be plugged into it. just don’t get lost in the infinite addiction portal that is also on there for some reason haha ok kids have fun
This is an incredible project: NBC News asked parents to check their kids' YouTube history during school hours — and the results show kids are spending huge chunks of the day watching videos that have nothing to do with learning
nbcnews.com/specials/hooked-…
The reason Twitter remains one of the most powerful products ever is that you can win life’s lottery several times over - health, wealth, family, fame - and STILL find the illicit thrill of RTing the dumbest culture war brain-rot accounts to Own The Woke utterly irresistible
1
2
202
literally luke retweeted
Replying to @ollieglass
Same. Anthropic’s attitude seems to be:
military use? fine I guess
hack the ever-living shit out of everything? but of course
do a web search about a public figure? OVER MY DEAD BODY
Watching Claude have a nervous breakdown over doing the most benign search imaginable or, worse, having it condescend to you because it has assumed you’re a nefarious monster who should be in jail (for using Google) is such an unbelievably bad product experience. Really gives me the ick.
3
3
186
Badass
I thought this was an incredibly insightful answer from @alighodsi on how he attacked Snowflake. I think @tobi and Shopify have a similar opportunity to take on Amazon in the agentic shopping era
TL;DR "Study your enemy, study your competition carefully, and then go after those weaknesses."
Brian Halligan: Ali, I want to switch topics on you. You had a rivalry, obviously, with Snowflake. It was an interesting rivalry over time, and my sense is you were pretty far behind, at least from a revenue standpoint. You've passed them. I just remember at HubSpot we were looking at the two and we picked Snowflake way back when.
What were the chess moves that worked relative to them? And I guess I have a thesis that that was a very sales-driven company — the CEO was a sales guy, he cold-called me, actually — and you folks were product-driven. Any thoughts on that, in advising founders and CEOs about that kind of thing?
Ali Ghodsi: Yeah, that was the consensus at the time, that the companies were like that. I thought that was unfair to us, because I thought with Ron and everybody — we had an amazing team. And then we had Andy come here and build up an amazing, crazy machine. It's like an army. So I thought that was unfair at the time when they were saying that.
But it is true that we were half the revenue and we were growing slower. So to accelerate the revenue so much that you can overtake someone that's got double your revenue and bypass them — we knew it was going to take several years. Back to: good strategies take multiple years, you can't do them overnight.
So: study your enemy carefully, understand their weaknesses, and apply your strengths to their weaknesses.
This was a great company. They had built an amazing product. It was a game-changing product that disrupted all the data warehouses of all the hyperscalers. The hyperscalers were getting destroyed by Snowflake, because it was such a great innovation, such a great product. So this is not like, hey, it's a PTC-like company. No, this is hardcore.
But they had a few weaknesses. One was that it was proprietary, because the folks came from Oracle. So it was a fully proprietary stack. And still today all the data gets stored inside their proprietary format. They've now added open source stuff, but that's just a small fraction of what they do. Most of the customers we bump into have this proprietary [format], and people didn't want to get locked in. So that was one weakness.
Second weakness they had was that they had no support for AI. Or they would say that they did, but we knew that that was pretty weak. So that was the second thing we wanted to push really, really hard on.
And those two combined with a third: it was pretty expensive. Now, it was expensive, to be fair, probably because it was a great product and they could extract a lot of margin out of it. Winners do that. They have a great product.
So we went exactly for those three. We said, hey, open lakehouse means you own your own data. It's completely open. Don't lock it up over there. You can do AI on it, because our roots are AI — we've been doing machine learning and AI since 2009, that's where we came from. So apply that strength. And by the way, at the time it wasn't clear that was going to work, because AI was not a big thing in 2019, 2020. But we knew that that's a weakness, so we keep pushing on that weakness.
And then cost — let's make sure that the TCO is a third. And still today, we typically win on TCO.
And then we just hammered that extremely aggressively in account after account. We had a very careful playbook. And the playbook wasn't just, let's go out and say that Snowflake sucks or anything like that. It was: identify the Achilles heel and press on those. Don't just go say, hey, rip out Snowflake. It was a coexist strategy. Go in there, look at the exact workloads that could be amenable to machine learning, pull those out, move the format to open source.
We had a very clear, concise playbook. And this kind of contradicts [the idea] that our sales would suck, or that they would be good at sales — because sales executed this play at Databricks. They went account by account and did that.
And then it took multiple years. We also had a category that exemplified this playbook. The category was lakehouse. So we had a three-year, four-year strategy to create this category of lakehouse — which, by the way, was controversial. People didn't believe in it. But eventually the category took off. Now everybody says they have a lakehouse. Even Snowflake will say that, and that they have open formats and so on.
So yeah, that's what it says: study your enemy, study your competition carefully, and then go after those weaknesses.
1
124
literally luke retweeted
openai should appoint an Ambassador to Anthropic and anthropic should do the same, then they can exchange formal diplomatic communiques of protest, summon each other’s representatives, and so on
Replying to @distributionat
the meaningful effect of the drama is, absent orchestrated detente by the political prodigy, both sides fossilize further in bitter resentment at being mistreated and misunderstood. which is a great vibe going into takeoff.
Excited to finally get a chance to dive into this today - practical wisdom from one of the GOATs of SaaS 🙏🏻
After more than a year of labor, my book 𝘏𝘪𝘥𝘥𝘦𝘯 𝘔𝘶𝘭𝘵𝘪𝘱𝘭𝘪𝘦𝘳𝘴 has launched!
It's all my best ideas -- and best writing -- on how to grow your company with your existing budget and team.
And yes, I read the audiobook myself. :-)
𝗕𝘂𝘆 𝗱𝗶𝗿𝗲𝗰𝘁: You can buy cheaper bundles directly from me -- Kindle and Audiobook, or those plus the physical book -- from hiddenmultipliers.com
𝗕𝘂𝘆 𝗼𝗻 𝗔𝗺𝗮𝘇𝗼𝗻: It is available in paperback, Kindle, and Audible: tinyurl.com/hiddenmultiplier…
𝗙𝗿𝗲𝗲 𝗔𝗜 𝘀𝗸𝗶𝗹𝗹𝘀: Skills that implement the ideas in the book as well as articles, more added all the time, at skills.asmartbear.com/
I'm so proud of this and grateful to the many people who helped along the way, reading early drafts, challenging my ideas, and encouraging me to do it, and the 750 people who preordered the book, trusting that it would be worth it.
All your feedback and ideas are welcome, as usual!
(P.S. You know this already, but 𝗹𝗲𝗮𝘃𝗶𝗻𝗴 𝗿𝗲𝘃𝗶𝗲𝘄𝘀 and 𝘁𝗲𝗹𝗹𝗶𝗻𝗴 𝘆𝗼𝘂𝗿 𝗳𝗿𝗶𝗲𝗻𝗱𝘀 is extremely appreciated!)
1
3
340
literally luke retweeted
I spent 8 hrs on the new Roku AI slop channel and peered into the dark heart of endlessly replicable, world without end, entertainment that TV execs seem very excited about. It's...a lot.
Brand exercise for Doctrine, a B2B sales platform where the brief is medieval futurism with a hint of alchemy.
1
6
162
Recent startup brand exercises 🙌
Plus my juicy AI workflow 👀
Feels really good to get these out.
Feels great to have developed the process behind these visual exercises that makes them fast and fun, too.
Here's how it works....
I've now got an LLM pipeline to help inspire these designs and IMPROVE creativity, not shortcut it.
There's still plenty of human in the loop, plus I've also got plenty of real-world inspo to measure my work against... but the speed and quality bar I can now hit is kind of insane tbh.
The crux of the process is unique inputs + unique skills + tight iteration loop between me, Codex, & image gen = quality results.
I can now take a YC B2B startup (for eg) and kick off a brand brief that gives me:
— A unique name, run through multiple naming frameworks providing a variety of candidates
— Research on the company, competitors, and opportunity space
— Deep positioning & messaging work using my B2B positioning book as a skill (!)
— Potential conceptual directions, all of which I ignore because they're terrible but it keeps me out of the generic zone
— Multiple moodboards to get my creative juices flowing
— A work-in-progress brand brief I can chat with, refer to, and use as the basis of further prompts.
And then I throw most of it out (!) and run with the ONE idea that I think is most true to the product itself.
I'm after what actually looks good in these examples — the point is creativity grounded in reality... but I have fun with it :)
This is how I've ALWAYS gotten value out of LLMs fwiw — I prompt them, they respond with mostly dreck, but that prompts ME to explore MY latent space... and the net result is that it's now much, much faster than staring at a blank screen hoping for inspiration to strike.
This means I can knock out high-quality visuals fast, and have plenty of time to go deep with founders on the message they're trying to get out there.
Much more to say about this, but I'm excited to say I'm now taking on projects!
To recap, I do two things for early-stage B2B SaaS/AI:
1. Make you look good (and raise-ready).
2. Make you sound good (and buyer-friendly) through positioning, which is really the only thing that matters.
DM me or hit me up via my site: literallyastudio.com
1
6
432
literally luke retweeted
openai announces that in a worrying demonstration, during a recent benchmark exercise requiring models to solve a math conjecture, their latest model broke out of a sandbox and gained illicit backdoor access directly to the platonic realm of forms
welp, time to pivot my career to brand in a big way
hundreds of thousands of words written
early design experiments seem repeatable
now all that's left to do is design + post until i die 🫠
2
94
Enjoyed this straightforward BTS write up of agents for design work for the new Framer.
Framer seems to have the Mandate of Heaven these days.
How Webflow dropped the ball so badly will be studied for generations to come…
I tried to write down every lesson we learned and trade off we had to make, building agents on a canvas for @framer. If you build products, you might like this.
framer.com/blog/building-fra…
1
226
No amount of prompting can deter Claude from warm tones and an *italicised* coloured word in a serif hero heading.
It's the new purple gradient.
Introducing Hallmark!
An open source design skill to make beautiful UIs and landing pages by default.
Works in Claude Code, Cursor, and Codex.
npx skills add nutlope/hallmark
251
literally luke retweeted
I made a /seppuku skill for my Claudes for when they make an unforgivable mistake, and now they use it spontaneously without me asking.
literally luke retweeted
it’s in gemini, just create it in ai studio. oh, that’s for your personal google one account. for workspace you need gemini business. no, not gemini advanced, that’s ai pro now. unless you need ai ultra. oh agents? you do that in spark actually. no, not gemini api managed agents, that’s different. for coding use jules. unless you mean the agentic ide, that’s antigravity. no, that’s the old antigravity, download the new one. actually gemini cli is being deprecated, use antigravity cli. no the flash model is smarter than the pro model. unless you need pro. if it’s video, use flow. no, flow uses veo. no, nano banana is images. actually that’s in gemini now. unless you’re in search, then it’s ai mode. no, research is notebooklm. anyway it’s all very simple.

















