Michael Williams

3.7K posts

Michael Williams banner
Michael Williams

Michael Williams

@mvwi

product + ai tooling + organizational design head of product @serotonin_hq · https://t.co/qrlpIurEp6 be the training data you want to see in the world

Katılım Ağustos 2013
837 Takip Edilen1.8K Takipçiler
Sabitlenmiş Tweet
Michael Williams
Michael Williams@mvwi·
If you're in product, learn engineering and design If you're in engineering, learn product and design If you're in design, learn engineering and product Learn marketing
Gokul Rajaram@gokulr

PREDICTION: THE CPO ROLE, AS WE KNOW IT, WILL VANISH IN FIVE YEARS At young AI native companies, the traditional PM role is on the wane, replaced with a product builder archetype that’s a combination of Product, Design and Engineering. These companies will never hire a CPO. A separate product leader leads to too much cognitive dissonance when the IC roles doing the actual work are blending, extra overhead and imposes an unnecessary coordination tax on the product development organization. Five years from now, these companies will be the leaders and set the cultural tone for the next generation, so my prediction is that all tech companies will stop hiring for the CPO role in five years. There will be a singular product development leader at each org. Ironically, this new role might still be called the CPO, except they will run the entire product development org. CPTO is far too unwieldy of a title and only exists today to alleviate confusion. Career implication: early / mid career product leaders need to stop aspiring to become CPOs. instead, you need to develop a panoply of product development skills across all three disciplines (+ analytics), be able to fluidly navigate the roles, and become a product builder, period. Farewell, CPO! It was a good 15-20 year run for this role in tech. But like everything else, it’s time to evolve.

English
0
0
7
459
Naruto11.eth
Naruto11.eth@naruto11eth·
this is what i have done most of my life. whenever i wanted to work somewhere, i have simply just dmed the founders and friends without thinking on what i'd contribute. the trick is to not make it a cold dm. you need to be providing a value that the company is missing. so look for what things you bring to the table. bring so much energy and value that they cant ignore you or find a way to fit you in even if they dont have that particular role for u. your resume wont be shortlisted because now recruiters get 100s of application every hour. it is hard for them to go thru everything even with ai. stand out by being high agency person. build things, improve their docs, work on their community, bring many things to the table.
Richard@Rodairos

Best advice I could give someone job searching: Skip the job application. Email the hiring manager, executive, founder, or CEO and send them something related to the job: >> Doing the actual work >> Share the strategy >> Show execution >> Send an artifact, codebase, etc Show them you can do the job and you'll skip to the front of the line. Applications are a waste of time. Getting shit done gets noticed.

English
13
1
126
8.7K
Michael Williams
tldr: write context files like a textbook - thin table of contents at the start - detailed chapters as reference material load the toc every session, load the chapters as needed llm brains were trained on humans -- the structures that make info usable for us work for them too
Garry Tan@garrytan

x.com/i/article/2017…

English
0
0
1
164
Michael Williams
@CalmCoding @ThePrimeagen mechanical enforcement is the only way make linters, typechecking, architecture tests, depcheck, knip, complexity checks, etc. blocking -- go a little overboard If the model *can* drift, it will
English
1
1
1
61
ben
ben@CalmCoding·
@ThePrimeagen The issue isn't really codifying rules, but rather the AI actually following them. Anyone have any tips on making the AI actually follow your rules?
English
5
0
0
3.4K
ThePrimeagen
ThePrimeagen@ThePrimeagen·
I am slowly coming around to AI assisted programming. I am genuinely trying to codify every rule about programming that I have and using that + several stages to build out small changes. Not sure the productivity changes, but I think I can see a modest gain in speed. I am also trying to be concerned about every line produced, not just slop trebucheting code over the wall.
English
387
62
3.5K
456.1K
Michael Williams
@ThePrimeagen I think the biggest unlock is moving to enforce quality at a mechanical level -- models are shit at staying consistent but with a small army of linters, architecture tests, depcheck, knip, etc. as buffers you can actually spit out pretty decent code at speed
English
0
0
0
49
Michael Williams
I'm all about people figuring out their own workflows but tbh if you aren't using voice to text w/ your agents then you're doing it wrong
English
0
0
1
49
Michael Williams
Michael Williams@mvwi·
@gregisenberg I think it's fair to say that machines are already using the internet far more than people The only thing that AI is changing is the amount of surface area we interact with (and our ability to compress the same intent into that smaller surface area)
English
0
0
0
45
GREG ISENBERG
GREG ISENBERG@gregisenberg·
Sometime in the next 2-3 years agents will be using the internet more than humans We designed the whole thing for human eyes, human emotions, human attention spans Agents do not have any of that The internet as we know it was built for the wrong user The opportunity is rebuilding everything for the new user Agent-native search. Agent-native commerce. Agent-native discovery Every category is open again I can't stop thinking about it.
English
291
100
1.1K
65.3K
Bryan Johnson
Bryan Johnson@bryan_johnson·
I put ice on my balls during sauna and my motile count went up 57%. Then I removed the ice and motility crashed 57%. 238 sessions to learn that the most important variable in my sauna protocol was a bag of ice on my groin. I am a serious scientist.
Bryan Johnson tweet media
English
499
226
9.6K
2.5M
Michael Williams
Michael Williams@mvwi·
@DrewAustin Is your main frustration with the 'crm collapse' knowledge input or knowledge retrieval? Those feel like very different problem spaces to me
English
1
0
0
212
Drew Austin
Drew Austin@DrewAustin·
I've been obsessed with the "second brain" concept for years. Professionally, I built and sold a company in agentic recruiting that touched on this. Personally, as someone with ADHD, I consume a massive amount of information but retaining and retrieving it is a nightmare. I've been life-logging for years, recording most of my Zoom conversations over the past 5 because I knew the tech would eventually catch up. Now it has. And I'm trying to bring this second brain to life. My vision: is sort of a Perplexity for my world. A personal knowledge network built from my conversations, interactions, content consumption, activities, relationships, and thoughts, that enables better retrieval, task completion, insight generation, and opportunity discovery. The challenge: as I start importing information, it immediately starts to feel like a bloated modern CRM. In my professional world as an investor, contacts, companies, and deals are key entities and important sources of knowledge. But the moment I ingest that structured data, CRM creep takes over and the "knowledge network" starts feeling like a database. Karpathy introduced an interesting model for this, but laser-focused around a single topic and research. I've been trying to take lessons from that approach and apply it to something broader and more personal. My open question: How do you structure a personal knowledge system that holds rich, structured entity data (people, companies, deals) without collapsing into a CRM? How do you keep it feeling like a living network of insight rather than a filing cabinet? Would love thoughts from anyone thinking about this problem.
English
28
3
54
7K
Michael Williams
Michael Williams@mvwi·
I've used ~$1100 in Opus/Sonnet API tokens over the last ~60 days Running this through Mythos instead would have cost about ~$7,850 (based on what we know re: pricing so far) $30-50k/mo in API credits per FT engineer seems likely at this rate
Michael Williams tweet media
English
0
0
1
387
Michael Williams
Michael Williams@mvwi·
Will be interesting to see if models this powerful make software practices slide back towards waterfall Not because agile doesn't work But because the one shot capability + level of expense in doing so make heavy upfront planning the economically rational way to build
Anthropic@AnthropicAI

Introducing Project Glasswing: an urgent initiative to help secure the world’s most critical software. It’s powered by our newest frontier model, Claude Mythos Preview, which can find software vulnerabilities better than all but the most skilled humans. anthropic.com/glasswing

English
0
0
1
75
Chintan Turakhia
Chintan Turakhia@chintanturakhia·
Lines of code count, then PR count, then Token count, then Agent count... what's next for Goodhart's Law?
English
8
0
13
916
jihad
jihad@jaesmail·
Do you know anyone that plays the LinkedIn game well? They’re getting distribution/driving leads but they don’t sound like hustle slop?
English
11
0
38
6.4K
Michael Williams
Michael Williams@mvwi·
creativity 4/10 audacity 8/10 utility 0/10 the sad part is how often this probably works
Michael Williams tweet mediaMichael Williams tweet mediaMichael Williams tweet media
English
0
0
0
43
Nic Barker
Nic Barker@nicbarkeragain·
@waim I hadn't seen this, thanks for letting me know!
English
1
0
5
1.6K
Nic Barker
Nic Barker@nicbarkeragain·
I'm completely convinced at this point that the "Command Palette" is a fundamental UI concept, and should be in all applications. It should also be a built in browser concept, there should be an API for websites to push items to the command palette ("new post", "muted words" etc)
Nic Barker tweet media
English
118
123
1.8K
198.6K