sanjaypillai

@docjamesw 👍🏾am traveling till Oct 30. If I can call or email now to buy then I’d like to do so. Else I’ll come by in person on Oct 31

@sanjaypillai Presale is in-person at Side Hustle Kirkland. Come on by and grab your tix. Online sales start Nov 1. Sold a bunch of tickets yesterday.

today i did two things i haven't done in a long time. drove to @microsoft (that was weird) and taught the art of stage presence. honored by the invitation to speak at my alma mater and humbled by how many people recognized and remembered me. now for the announcement ...

@docjamesw In

the event will be private and strictly limited to 50 people. i will consider live-streaming if i can get a sponsor to sign up for the broadcast. who's in?

@buccocapital The Snow Leopard - Peter Matthiessen. Such a gentle but powerful book. I think I read it at the right time for me.

Let’s crowdsource this: What’s the last really great book you read? Can be fiction or non-fiction.

…to problems - to set aside preconceived notions and focus on WAYRTTD. Part of the benefit for me is stepping away from day to day work to refocus on the art of product and critically examine whether I’ve fallen prey to lazy thinking and other bad shortcuts. 2/2

Enjoyed spending time with a great group of senior PMs & PM leaders in @shreyas new product sense course. His direct, no-nonsense style is much appreciated as is his constant reminder to focus on clarity of thinking and trusting in one’s own ability to come up with answers… 1/2

Enjoy this playlist (with 96 videos & counting) anytime you’re looking to watch or listen to (or binge on 😄) advanced product & leadership content. (this link will open the playlist in the YT app, so you can save the playlist for later) yt.openinapp.co/shreyas-playli…

Great thread by @mipsytipsy on power symmetry in organizations. I could get “never confuse lack of clarity with empowerment” tattooed on my forehead. One of the most impt jobs of managers and leaders is to internalize purpose & communicate it clearly. Direction ≠ disempowerment.

Fighter planes of the 1950s used the Bendix Central Air Data Computer, an electromechanical analog computer that computed with gears. In this thread, I look inside its pressure transducers that converted pressure readings to shaft rotations that turned the gears. 1/13

lots of ppl were asking, why not an app. here’s my personal opinions on a thread 🧵:

@dieworkwear Also Glasswing Seattle glasswingshop.com/collections/ka…

Some things to check out: 1. Ring Coat: Originally designed in the early 2000s, this is one of the company's flagship pieces. They do them in every material imaginable. It's something like a Japanese kimono fused with a US Army field jacket. Massive hood, plenty of pockets, and kimono sleeves. It comes in short and long lengths (I prefer the short version). Button it all the way up for laughs. 2. Giant Scarves: Kapital also does these huge, felted scarves with large motifs (including a few with cats, for fellow cat lovers). The only catch is that the size and thickness of these scarves give them a lot of volume, and I find they go best with big coats in Kapital's range, not stuff like trim Italian topcoats. You need something with a bit of room for layering. 3. Funky Pants: The company also makes a bunch of funky pants every season. I find the double-knee carpenter pants easiest to wear, but every pair is a joy to put on. There's a great video of @modsiwW wearing their flared-legged jeans with a bunch of hippie patches on them. As always, he looks incredible. Some channel the John Lennon vibe in the photo above that inspired this post. 4. Navy Supply Boots: Finally, let's end with something supremely easy to wear. These side-zip boots come in various types of leather. The brown ones pictured below were dyed in persimmon juice (how cool is that?), which allows them to fade in really unique, antique-ish ways. You can wear these with all sorts of stuff from Kapital, or just a very basic pair of slim-straight jeans worn with a flannel shirt and a trucker jacket. Slots into any kind of workwear wardrobe. They're a bit hard to find in the US, but you can place a special order through Cotton Sheep in San Francisco (they were the first US stockist for Kapital's line). You can find Kapital at Standard & Strange, Canoe Club, Blue Button Shop, Mr. Porter, Blue in Green Soho, Mannahatta NYC, and Union Los Angeles. The line is kinda funky, so buy from a place that allows for easy returns.

All I'm saying is that if you love this aesthetic, you will love Kapital, a Japanese brand that riffs on workwear, Americana, and folk clothing. The brand has taken a bit of a hyped streetwear turn in the last few years, but it has deep roots 🧵

Superblock concepts can be easily applied to American cities. Here's one I proposed for the Capitol Hill neighborhood in Seattle. x.com/Qagggy/status/…

Yes. Being a great engineering manager who cares about people is exhausting. Professionalism itself is a form of mild dehumanization; you and I set our identities and personal selves to the side, and interact with each other's professional functional self. We become cogs. 🙃

I am the first author of the Galactica paper and have been quiet about it for a year. Maybe I will write a blog post talking about what actually happened, but if you want the TLDR: 1. Galactica was a base model trained on scientific literature and modalities. 2. We approached it with a number of hypotheses about data quality, reasoning, scientific modalities, LLM training, that hadn’t been covered in the literature - you can read about these in the paper. 3. For its time, it was a good model for its domain; outperforming PaLM and Chinchilla with 10x and 2x less compute. 4. We did this with a 8 person team which is an order of magnitude fewer people than other LLM teams at the time. 5. We were overstretched and lost situational awareness at launch by releasing demo of a *base model* without checks. We were aware of what potential criticisms would be, but we lost sight of the obvious in the workload we were under. 6. One of the considerations for a demo was we wanted to understand the distribution of scientific queries that people would use for LLMs (useful for instruction tuning and RLHF). Obviously this was a free goal we gave to journalists who instead queried it outside its domain. But yes we should have known better. 7. We had a “good faith” assumption that we’d share the base model, warts and all, with four disclaimers about hallucinations on the demo - so people could see what it could do (openness). Again, obviously this didn’t work. 8. A mistake on our part that didn’t help was people treated the site like a *product*. We put our vision etc on the site, which misled about expectations. We definitely did not view it as a product! It was a base model demo. 9. Pretty much every LLM researcher I’ve talked to (including at ICML recently) was complimentary about the strength of the research, which was sadly overshadowed by the demo drama - yes this was our fault for allowing this to happen. 10. Fortunately most of the lessons and work went into LLaMA 2; the RLHF research you see in that paper is from the Galactica team. Further research coming soon that should be interesting. It’s a bit of a riddle because on the one hand the demo drama could have been avoided by us, but at the same time the “fake science” fears were very ridiculous and despite being on HuggingFace for a year, the model hasn’t caused any damage. To reiterate: the anti-Galactica commentary was really stupid, however we should not have allowed that to even happen if we had launched it better. I stick by the research completely - and even the demo decision, which was unprecedented openness for a big company with an LLM at the time, wasn’t inherently bad - but it was just misguided given the attack vectors it opened for us. Despite all the above, I would do it all again in a heartbeat. Better to do something and regret, then not do anything at all. Still hurts though! 🙂

🤬USERS ARE IDIOTS IGNORING YOUR MESSAGES, BLAMING YOU?🤬 Here's the thing about communication. After a few words your possibility space is more atoms than the galaxy. You haven't iterated enough. You need to experiment. Here's what I've learned ~20 yrs in support messaging.🧵

In my opinion, this is really what differentiate a Principal from a Staff and a Staff from a Senior Engineer. Softskill-wise, there’s no better predictable of someone’s maturity level and capacity to influence an organization at scale & to grow than those 3 macro-skills: 1. Sponsorship (truly supporting other people's ideas), 2. Selfless & Egoless way of leading (Don’t be a politician: Say “I don’t know” & admitting your mistakes) 3. Be Open to Influence (To lead, you have to follow. To influence, you have to be influenced) I know, I know, what about accountability and ownership? What about the ability to drive things and make them happen quickly? What about the ability of distilling complex and ambiguous problems into a simple solution? What about coaching and mentoring? What about the super deep expertise? What about the ability to navigate and coordinate efforts across the org to deliver big cross-area projects? What about thinking strategically and planning for the long-term? These are super important too, they are crucial. But, in my experience, they are not the best predictable of what is going to make someone succeed and scale their impact and their career beyond the Staff Level. Confidence to sponsor other folks ideas even when they are super shiny or bold ideas and even when you would probably solve slightly differently without “cookie leaking” and without getting credits to yourself. Selfless and egoless way of leading. You are an engineer, You are not a politician. Admit your mistakes and what you got wrong as a Senior Engineer, your team will respect you a lot more (and be a lot more influenced by what you share) in the long run. Give your support quickly to other leaders who are working to make improvements. Even if you disagree with their initial approach, someone trustworthy leading a project will almost always get to a good outcome. If there's something you disagree with but only in a minor way, let others take the lead figuring it out. A helpful question here is, "Will what we do here matter to me in six months?" If it won't, take the opportunity to follow. Great questions to ask: - Tell me the last time you supported one of your more junior engineers ideas - Tell me about the last time you admitted a mistake you made to your team - Tell me about the worst technical mistake you made over the last couple of years as a Staff in the company - Tell me about the last time someone on your team changed your mind

@johncutlefish I do this too so I’m glad to say there is a term for it. Omne trium perfectum.