Chief AI & Co-founder @AnacondaInc; invented @pyscript_dev, @PyData @Bokeh @Datashader. Former physicist. A student of the human condition. bsky: @wang.social

Austin, tx
THE MOMENT YOU'VE BEEN WAITING FOR!! Type "=PY(" into Excel, and start executing Python directly in the @msexcel grid! Really excited about our new partnership with @Microsoft to democratize data science, machine learning, and AI to all knowledge workers!
We’re excited to unveil Python in Excel! Get ready for a whole new way to execute advanced analytics capabilities from within Excel 🐍 + 📊 = 💚 Check out the new integration btwn @anacondainc & @msexcel, @Microsoft365 here 👇 bit.ly/3KSblQ6
5
39
190
49,289
Um... I just had like a 20 minute conversation with ChatGPT about the history of modern physics. If I had this shit as a tutor during high school and college.... OMG. I think we can basically re-invent the concept of education at scale. College as we know it will cease to exist.
606
2,214
18,443
"Shakespeare, but with zombies, dragons, and an odd fixation with castration."
45
572
4,731
When humanity does create AGI, it will be named Untitled14.ipynb
41
261
2,435
220,364
Fine, I’ll bite. I LOVE to code. I love creating good abstractions that model interesting problems. I love writing the help text for little command line tools, channeling the craft of the Great Unix Gods of Yore who wrote the tools we all use and enjoy. As a CEO, I miss coding.
24
51
1,140
264,851
I'm excited to announce that I am stepping into the role of CEO at Anaconda. Blog post here: anaconda.com/anaconda-enters… Thanks to all our customers and users, and looking forward to the road ahead! @anacondainc
91
87
1,024
Had a great time with @lexfridman talking about life, the universe, and everything, including of course Python & computers and all that. 😀
Here's my conversation with Peter Wang (@pwang), one of the most impactful people in the Python community. Also, he is a physicist, philosopher & someone who both @ericweinstein & @teoliphant said I absolutely must talk to, so I did and loved it! piped.video/watch?v=X0-SXS6z…
30
62
999
This is the real danger of “AI”: that even when it’s pure snake oil, there will be a tech hype machine eager to foist half-baked nonsense onto the world. Normal humans & policymakers are easily dazzled by aesthetics of the presentation, and can’t begin to question the substance.
👓 Facial Recognition is able to analyze in real time the Emotions & feelings! #AI #MachineLearning #FacialRecognition 🎥emotion @DeepLearn007 @PawlowskiMario @mvollmer1 @gvalan @ipfconline1 @diioannid @ShiCooks @kalydeoo @Ym78200 @Nicochan33 @chboursin
28
200
841
"Instead of Matlab, everyone uses Python now" -- CEO of @nvidia, name dropping @teoliphant @wesmckinn et al
11
343
836
Sigh.. ok I’ll explain this one more time: Voting is not a great way to surface the “best ideas” from a group. The primary purpose of voting is to maintain legitimacy of the regime in the minds of the losing cadre; you get their “buy in” bc they were part of the “process”
28
81
755
117,477
I'm dying 😂😂
6
271
710
Also in 2006: @teoliphant released the first version of Numpy, unifying the array libraries in Python-land and starting a multi-decade revolution in scalable, vectorized computing for everyone. It paved the way for modern machine learning and AI.
8
104
636
LLMs are actually as groundbreaking as eg electricity and internal combustion; but for them to manifest their full potential, it will require the deconstruction of almost all the institutions of modern industrialized society, because they require an expansion of our concept of *what is human*. Neither the AGI Alarmists nor the AGI Skeptics are thinking big enough. AGI Alarmists think we might be able to preserve a lot of existing institutions & concepts if we Try Hard Enough; Skeptics want to believe we have more time, to Figure It All Out. We are already two steps down the road into an unknown world, and that path almost certainly will be paved in blood, which almost certainly implies the rise of higher-coherence institutions that current philosophies can only (poorly) perceive as “authoritarian”. The Revolution will not be pre-authorized.
aii, let's do this... AI/ML version what’s your critical hot take on AI/ML that would have you in this position
38
109
677
420,725
💯 Smaller models, tuned and configured to reason, coupled with RAG and knowledge graphs, will absolutely eat a lot of white collar jobs. Maybe not in the next 6 months, but it’s inevitable. That’s why it’s so important that the open source / open LLM community start *now* on standards for data governance, provenance, and licensing. We have the biggest opportunity for democratizing and resetting the scales of economic equity, since the rise of consumer internet and seminal open source like Linux and GCC.
You thought that you can go to sleep now?? Orca 2 Just dropped. Paper: arxiv.org/pdf/2311.11045.pdf Results: Orca 2 13B beats LLaMA-Chat-70B TL;DR: Training smaller model to reason by using multiple techniques: step-by-step, recall then generate, recall-reason-generate, direct answer And determining the most effective solution strategy for each task.
15
109
668
277,153
My dear mother passed away unexpectedly a few days ago, and we buried her Monday. Here is my eulogy, with photos, in a blog post. Even though it's long, it's not long enough to do justice to the amazing life of a most extraordinary woman. 😢❤️ medium.com/@pwang/goodbye-mo…
98
27
623
That sure looks like a Matplotlib window :-) @tacaswell
Congratulations to Katie Bouman to whom we owe the first photograph of a black hole ever. Not seeing her name circulate nearly enough in the press. Amazing work. And here’s to more women in science (getting their credit and being remembered in history) 💥🔥☄️
12
104
612
OH SNAP!!!!! #Python just hit #1 on TIOBE, displacing C and Java! Congrats @gvanrossum and the many, many contributors who made this happen! @ThePSF @PyData Tag a Pythonista in this thread! tiobe.com/tiobe-index/
22
132
566
Given the interest in PyScript and some of the common questions that are coming up, I figured I'd share some of my slides which may answer some questions. Also I highly recommend checking out our engineering blog post on it: engineering.anaconda.com/
8
135
535
I'm an amateur astronomer since the age of 12. If any company shits all over the sanctity & cosmic beauty of the night sky by deploying advertising cubesats, I will make it a life mission of mine to destroy each and every single one - either in space or on the launchpad.
21
109
473
Missing from the narrative: Hundreds of thousands of open source Python devs, and ML reseachers publishing their work. The AI Revolution built on the backs of a *community* innovation effort in Python, DS & ML. If we want to keep it open, we have to stand up and be counted 💪
The AI race: Microsoft + OpenAI Google + DeepMind Tesla + xAI Meta + Meta AI and now Amazon + Anthropic History is unfolding in front of our eyes. The AI boom will be bigger than the internet.
15
91
485
86,403
This is going to be a major Tshirt trend over the next couple of years: Innocent-looking shirts with various patterns that are specifically designed to trick neural networks. Adversarial hats will also be a thing, for face detectors.
We're living in a cyberpunk future: “Fooling automated surveillance cameras: adversarial patches to attack person detection” arxiv.org/abs/1904.08653
12
189
445
The fundamental law of data science & AI: "Your predictions are only as good as your data." For every fancy algorithm you devise, there's some dude with a wagon full of phones running game on you. lnkd.in/eHmQZjV
6
144
416
Amazing and terrifying. You can take one photo of someone, and then make videos of them in arbitrary facial poses. Can't see any ways this could go wrong!
Samsung researchers have released a model that can generate faces in new poses from just a single image/frame (for each of face, pose). Done by building a well-trained landmark model in advance & one-shotting from that. This gif is my favourite... techcrunch.com/2019/05/22/mo…
7
204
413
8 yr old: "Is it octopuses, or octopi?" Me: "It's octopodes, but octopuses is also acceptable. Because 'octopus' is Greek and not Latin,..." 8 yr old: "Never mind, I did not want to have this conversation right now."
5
49
411
This slide from #AnacondaCON might be of interest to some... 😁🍉🍉🌮
14
118
417
📢📢 Since people seem to be quite confused about this...To be extra, super duper clear: PYSCRIPT RUNS IN THE BROWSER VIA WEBASSEMBLY IT IS NO LESS INTRINSICALLY INSECURE THAN YOUR JAVASCRIPT CODE IF YOU HAVE AN ISSUE WITH THAT, TAKE IT UP WITH YOUR BROWSER VENDOR Thank you.
6
45
394
Big news! I'm proud to announce the @AnacondaInc OSS Dividend! We're committing 10% of revenue from our newly-launched Commercial Edition to @NumFOCUS (through end of this year). In 2021, we're committing 1% of all revenue. Viva #python #opensource! anaconda.com/blog/sustaining…
11
88
365
OMFG
The team at @Deezer just released #Spleeter, a Python music source separation library with state-of-the-art pre-trained models! 🎶✨ Straight from command line, you can extract voice, piano, drums... from any music track! Uses @TensorFlow and #Keras. github.com/deezer/spleeter
10
52
360
The frustration and pain of Python packaging comes from two major factors: 1) it’s an open source ecosystem (with HUGELY diverging user personas), and all scaled open source ecosystems end up settling on the concept of a “distribution” to solve the n^2 coordination problem between projects. Every time you install something from PyPI, the UX deceives you into thinking you are doing something like rpm or apt. But what you’re actually doing is rolling your own Python distro from scratch, as if you were building a new Linux distro from sources. Most people don’t want to be (and are not qualified to be) their own *distro* maintainer, and they don’t realize until much later the full consequences of taking on this task. (Eg This thread shows how quickly the complexity can spiral out of control - github.com/pypa/packaging-pr…) 2) The Python library ecosystem consists of a far, far gnarlier and more numerous set of native C/C++ dependencies than most other languages. This is also the reason for its succeed in data/ML/AI. Additionally, these libraries are often written by people who are not professional software engineers, and are oftentimes used by people who are completely new to programming. (You’d be shocked at the % of Jupyter Notebook users who don’t know what a “command line” is.) Node/JS is very seldom dealing with eg Fortran on Windows. I don’t know the last time someone tried to get raw device access on a raspberry pi from PHP the way that embedded folks do with PyVESA. Etc. I believe the Python packaging problem is tractable but it requires re-thinking the paradigms that the core dev community has been championing for quite some time (“let a thousand tools flourish”; “we don’t want to dictate a default” even though there is a de facto default; “scientific and numeric libraries are too weird so maybe let the conda zealots figure that out”.) I know everyone means well but they are mostly volunteers and the scope of the task has now (IMO) far exceeded what a ragtag group of volunteers can handle. I’m on a little bit of a mission this year to try to at least orient the conversation in the right frames, so that folks at the center of all this (across many different tools) can discuss the full set of inter-related concerns with a holistic perspective, and make the right product and engineering tradeoffs.
17
56
378
130,659
July 2017 is when "Anaconda (Python distribution)" started exceeding popularity of "Anaconda (Snake)" in Google rankings.
6
52
328
It’s really interesting how many responses to @pyscript_dev have been: “Why don’t people just learn Javascript?” I’ll wager many of these people could score a $200k salary bump if they just learned stats, vector calc, etc. to become an AI engineer. Why don’t they?
16
27
343
Panel: Easy interactive dashboards for Python ✅Bokeh ✅Datashader ✅Matplotlib ✅Plotly ✅Altair ✅Jupyter ✅Standalone deployment ✅BSD licensed open source Great job to the whole team!
We are very happy to announce the formal release of Panel: A high-level app and dashboarding solution for Python from @anacondainc. Many thanks to the contributors so far and to @BokehPlots for providing a great tool to build on. See the Medium blog post: medium.com/@philipp.jfr/pane…
2
59
292
I agree with Paul. I’ve been into computers since the late 1980s and could definitely see certain kinds of trends, and make useful predictions. AI (even non-AGI) hits different. It could very well go “foom” and quickly lead to 2nd and 3rd reactions in culture and politics. Humans generally don’t respond to change; we respond to changes in the *rate* of change. And when even that starts to become nonlinear, people start freaking out. The majority of people need a stable narrative for their lives, i.e. a linear first derivative. From what I can tell, we’re going to have (at least) a few decades of nonlinear first derivatives.
One of the most interesting (and also frightening) things about AI is how difficult it is to predict what will happen. My kids ask me what's going to happen, and all I can say is that there will be huge changes, and I can't predict what they'll be.
10
35
279
124,504
An example of how stupidity metastasizes: That stupid paper benchmarking "carbon footprint" of different languages has been making the rounds, and now we have had a job candidate ask us what are we doing "to make Python greener". There are no emojis to express how I feel.
24
44
260
Given that #Python 3.11 is almost out, has there been any discussion about fun Easter egg/gags to do for Python v3.14.159? Maybe call it Pithon?
8
36
256
The video of my PyCon keynote is up! This provides much more context around *why* we made PyScript and who it's for. TL;DR: it's about democratizing computing, data & quantitative computing, and the future of a hackable, accessible web. #Python anaconda.cloud/pyscript-pyco…
5
74
255
Me: "I'm glad Minecraft has a scripting API now!" 8 yr old: "Right. So now you can mod it with Python and not whine about having to write Java." #HeLearnedItByWatchingDad
2
34
247
(Tom Lehrer voice) "Don't write math formulas on stickers if you can't spell... them..."
8
83
261
Ugh, your Baphomet is missing boobs. A critical aspect of Baphomet is hermaphroditism
14
2
227
Gem from Reddit
7
79
247
“Kubernetes is the C++ of orchestration software. Immensely powerful, includes all the features, looks deceptively simple, and will hurt you repeatedly until you join its priesthood and devote your life to its mysteries.” Just read this again and it’s so hilarious and good.
2
38
250
May 12, 2022: It actually happened. I didn't have numpy installed yet on my local machine, so while I was waiting for a new env to be created and installed, I opened a browser tab and launched a local PyScript REPL, to do a quick bit of math. The end is near....
9
16
231
Here's a cool disruptive startup idea: A bank where you put your money in it, aaaaand.... it just stays there. And then when you want to get your money out, you can go get it out. It'll change the world!
35
15
235
52,473
Apple in 2019: My spacebar double-spaces constantly. I accidentally brush the "Esc" button on . the Touch . Bar several times a day. The "Pause" button in the Touch Bar doesn't pause iTunes. Once a day I have to run back to my office to get a dongle. This isn't . progress.
18
28
226
A year ago, I had the great pleasure of chatting with Lex Fridman for a few hours about philosophy, computing, the socioeconomics of open source, and many other interesting topics. 😀
Here's my conversation with Peter Wang (@pwang), one of the most impactful people in the Python community. Also, he is a physicist, philosopher & someone who both @ericweinstein & @teoliphant said I absolutely must talk to, so I did and loved it! piped.video/watch?v=X0-SXS6z…
12
14
236
66,221
I gave my son some money and a Robinhood account for his birthday a month ago. Despite my exhortations, he procrastinated on investing any of it and left it in cash. Now he’s up 13% over SPY. Follow me for more great passive investing advice
7
6
224
New rule: a C*O is not allowed to use the term "AI" unless *every* business/data analyst reporting to them can write the formula for a linear regression. Let's start with that.
14
46
217
I get the “operating system” metaphor but the key thing about operating systems were that they bridged different layers of abstractions, to reify a higher-order plane of things (“software”) above a lower plane of hardware voltages comprising bits. A lot of the immediate near-term lift from LLMs will be in orchestration and melting through a lot of accumulated cruft and tech complexity, and increasing the TAM of who can build apps with computers. But is there a higher level of abstraction? LLMs seem to be able to terminate directly into a human’s mental frame… no abstraction needed. To me, it’s entirely possible that we are crossing the threshold into a new kind of very bespoke information system architecture. The tough question for businesses and investors is: what if there really is no new durable, defensible, scalable monetization in most of this?
With many 🧩 dropping recently, a more complete picture is emerging of LLMs not as a chatbot, but the kernel process of a new Operating System. E.g. today it orchestrates: - Input & Output across modalities (text, audio, vision) - Code interpreter, ability to write & run programs - Browser / internet access - Embeddings database for files and internal memory storage & retrieval A lot of computing concepts carry over. Currently we have single-threaded execution running at ~10Hz (tok/s) and enjoy looking at the assembly-level execution traces stream by. Concepts from computer security carry over, with attacks, defenses and emerging vulnerabilities. I also like the nearest neighbor analogy of "Operating System" because the industry is starting to shape up similar: Windows, OS X, and Linux <-> GPT, PaLM, Claude, and Llama/Mistral(?:)). An OS comes with default apps but has an app store. Most apps can be adapted to multiple platforms. TLDR looking at LLMs as chatbots is the same as looking at early computers as calculators. We're seeing an emergence of a whole new computing paradigm, and it is very early.
10
31
220
102,997
Ok software industry: I'm reading some bios of senior pilots and the contrast between "coders" & aviation professionals could not be more stark. That entire industry has a strong sense of responsibility towards safety & operational excellence. Meanwhile we curl | sudo bash ✈️🤷‍♂️
10
60
209
Wrong on so many counts. Nuclear has killed orders of magnitude fewer people than coal. Three mile island: 0 fatalities. Chernobyl: 31. Fukushima: 0 However, popular hysteria about nuclear meltdown lead to gross over-regulation of the industry, which means less innovation and fewer nuclear plants, which meant more coal. Pollution from coal power kills ~3,000 people annually. piped.video/4c5RPk8FlIk @i_sodope_
Let's make #AI like biotech, where companies must demonstrate safety, rather than the civilian nuclear industry, where poor safety standards gave us Three Mile Island, Chernobyl, Fukushima and a backlash that crushed the industry: bnnbloomberg.ca/video/push-f…
25
21
203
53,553
We are at this stage of "AI". Just keep that in mind as you board pressurized Mach 0.9 passengers jets in a few weeks to see your family for the holidays.
5
29
206
Oldie but goodie
3
32
203
13,772
Yes, please tell me more about how this was a failure of algorithms and machine learning...
10
26
206
Elon’s retweet of this DeepFake video has now been viewed 110 Million times. That’s equivalent to the viewership of the Superbowl. Libel and slander laws exist for a reason. It’s time to use them on viral platforms that massively amplify disinformation.
This is amazing 😂
138
15
186
79,356
Founders become founders, in order to do things not to “be a founder”
What did he mean by this
17
12
198
52,301
As a coder, software consultant, and tech executive... I have seen many dumpster fires. But today I saw my first actual real life dumpster fire. With an actual dumpster. And no Javacript anywhere in sight.
7
21
183
As multiple crises pop off, we might be finally crossing the threshold into liminality. This is not a good thing! This is not some hippie Age-of-Aquarius bullshit. Liminality means people are on their own when it comes to sense-making. COVID was our “gear check” as a species and we failed horribly. Liminal zones for individuals is when new internal subjective states can be accessed, because you *have to* depart old ones. You *cannot* stay the same. It’s sink-or-swim. Liminal zones for societies are when revolutions happen. Old ways of sensing and cohering collective truth evaporate, and new cults of awareness appear, and then fight for dominance. Yes, fight. All war is culture war; in a liminal zone, it falls back to an even earlier state: religious war. It is quite likely that our *only* hope through this is to build an LLM/AI/AGI to facilitate collective sense-making which out-coheres the accelerating pace of liminal events which split us further into microtribes of confusion and delusion. We don’t need more podcasters “questioning the MSM narrative”. The narrative is crumbling all on its own. But we absolutely don’t want nuclear-capable states with no central narrative. Every society is like a school of disparate fish, eating each other but swimming together. Moloch is the stigmergy of the school itself. Boulders of change are now raining down into the ocean of our collective objective reality, and Moloch evaporates as all the fishes and dolphins and whales scramble to form new schools. The Israel-Hamas conflict will be a source of many boulders of confusion. Next year’s election will be a storm. My thesis is: whoever or whatever culture can most quickly leverage AI tech and a comms platform like Twitter, Bluesky, or Mastodon to support a *coherent* polyphony of subcultures of sense making, will become the operating system of a new humanity. The defining characteristic will be that it *relies* on AI to exist, just our current industrial world relies of electricity and literacy in order to exist. I think this should be the near-term focus of the e/acc movement. Who is actually *actually* working on this right now? (Not just gesturing vaguely about it while high at a burn) What projects and resources can we harness? @jim_rutt & @jgreenhall is it time for “rally point gamma”? 😉 CC @tristanharris @aza @edelwax @vgr @generativist @ericweinstein @pfrazee @arcalinea @creon @gregosuri @ellie__hain @brycehidy @SamoBurja @pmarca @danielschmach @lukenosek @balajis @chaosprime @richdecibels @cognazor @brett_fujioka @theannagat @visakanv @indian_bronson @espacewalk @roon @lydialaurenson @jposhaughnessy @mbrendan1 @JustJamieJoyce
50
30
192
56,410
“Instead of writing C extensions, I strongly encourage folks to use one of these other tools...” @teoliphant at @pydata Austin keynote
5
39
185
Programming will never advance, so long as we continue to view programming, debugging, and execution as separate activities. Alternatively: Compilers considered harmful. Python won via the REPL, but Excel is still king of accessible modeling. (h/t @rsnous for asking a good Q)
18
32
183
Had a random shower thought, and checked just now. Oumuamua’s direction of origin is Vega. Tsar Bomba’s detonation would have been visible from the direction of Vega. It was at 11:30 Moscow time in October, at 70° latitude. In the local sky, Vega was approx 80° in azimuth from the sun, and at 45° altitude. The sun was barely on the horizon (Arctic winter). So an observer in the direction of Vega would be fairly well-placed to observe a sharp spike in luminosity; it would neither be lost in the bright reflection of the sun (if the distant observer was looking at the earth face-on), nor lost in the glare (if the observer was directly behind earth). cc @creon 😉
3
20
187
34,010
As long as the Python community thinks of their problems as “Python packaging problems”, and not “legacy C compiler & linker issues from the 1970s which we inherited as the cursed prize of becoming the world’s most popular language for data and numerics”, then we won’t be able to solve it. Python packaging is a little messy… but it’s C/C++ packaging that’s the dumpster fire that we inherited. It’s not really worth comparing to any other language ecosystem that hasn’t inherited the same graphite dumpster fire. 🤷🏻‍♂️
So here is a question for you all? Should Rye exist? github.com/mitsuhiko/rye/dis…
13
28
185
98,569
This. This. THIS IS HOW IT'S DONE. Documentation-driven development is the approach I find most natural. If I can't explain it in the docstring, for a future user (or dev or maintainer), then I need to really rethink the design.
I find it really enjoyable to: ➕ Write out all the function signatures with no implementation ➕ Then add comments about what each function will do ➕ And finally, have smooth frictionless implementation because you've already answered all the questions during previous steps
6
35
171
Replying to @DoktorSly
🤷🏻‍♂️ We’ll see. So many replies to my tweet are complaining about the imperfections of a piece of software that was announced less than a week ago. It’s based on transformers, which was pioneered 4 years ago. Care to extrapolate progress over the next 10 years?
16
3
170
Replying to @ericweinstein
The WTFness of the Trump years caused many on the Left to shrink the liminal gap between Disagreement and Transgression, to nearly zero on several areas. They mistake a call for epistemic humility on eg critical theory to be “tolerance of intolerance”, and shut down the convo.
8
14
174
I, Peter Wang, user of @Apple products since 1985 and user of Mac OS X since 10.2, do solemnly swear to applaud Apple adding 4mm of thickness to the 2020 Macbook Pro refresh in order to restore the 2015-era keyboard, USB3 ports, and an HDMI port. PLEEEAAAASE?
9
20
168
I cannot express in words how much I HATE the narrative that "a single company will rule the age of AI". (and to be clear, this is nothing against Pedro himself; he just happens to be falling into the same narrative trap I see others falling for) AI is literally a transformation of our species. Companies and corporate structures are 17th-century constructs based on pre-industrial models of allocation of scarcity in the face of risk. The future of humanity is one of physical abundance for our biological forms, but UNLIMITED UPSIDE for our cognitive and spiritual essences. I refuse to submit my & my childrens' future to the lazy inertia of existing economic paradigms. We failed to secure open innovation in software, and we failed to lock the Web open. Instead we produced trillion dollar companies that held our data captive to create surveillance and attention economies, rather than empowering end users and fostering the development of end-user creativity and innovation. Two generations of investors are now completely hooked on this model of extractive network effects. SO MUCH of this growth was fueled by the gift economics of nerd communities producing open-source software and cohering de facto OPEN STANDARDS which neither governments nor BigCo nor academia could produce. These Schelling points that balanced risk, utility, and user needs are brilliant examples of alternative economic models for an abundant, regenerative future. And yet, just a few years into the AI era, we are allowing the narrative to be hijacked by investors, tech journalism fluffers, and an entire hype engine, to set up a paradigm and frame that "only the biggest XYZ" can win. This entire landscape is built on human ecologies that were the OPPOSITE of winner-take-all: thousands of unsung heroes in open collaboration, tirelessly hacking on code late at night and on weekends. And now a portion of the tech ecosystem - profiteers and marketers - have the panache to attack "open source". WAKE UP NERDS. The future can be abundant and all you need is a laptop, power, and ping. We can't let the narrative get hijacked. It doesn't have to be "winner take all", and looking to create "winner take all" dynamics is exactly what will lead to suboptimal configurations of the possibility space. cc @teoliphant @martin_casado @balajis @chacon @lexfridman @ericweinstein @jim_rutt
Google is a disaster, Microsoft is dependent on OpenAI, OpenAI is a tangle: the race to be the company that rules the age of AI is wide open.
19
49
181
43,157
New library! scikit-prettyleaf, an image classifier for autumn leaves, based on the aesthetic preferences of my 6 yr old daughter.
7
8
179
Replying to @generativist
Oh oh do this on a red eye on a Sunday night during daylight savings changeover
2
176
7,827
Oldie but goodie
2
20
177
It’s not a tool for collective sense-making; it’s a tool for popular control. Always has been, always will. If, somehow, some general sense of “the collective will of the people” can emerge from a voting process, it’s a nice side benefit.
6
7
179
10,307
There are only two hard problems in distributed systems: 2. Exactly-once delivery 1. Guaranteed order of messages 2. Exactly-once delivery
4
216
160
Fun fact: web browsers and servers are still not free. We just pay through other forms of taxation. Google's quarterly profits are $30 Billion.
Fun fact: web browsers used to be commercial products. You had to buy them. #Web30
4
70
158
Plots of race in #NewYork #Chicago & #LosAngeles using @BokehPlots and DataShader; just a few lines of #Python!
9
59
161
My father came here in 1979 with no money, speaking almost no English. He kept some of his notebooks from that time. In reading them years later, I saw the huge effort he put into learning the language & American culture. Fuck your anti-immigrant bullshit @FoxNews @TomiLahren
.@TomiLahren: "You don't just come into this country with low skills, low education, not understanding the language and come into our country because someone says it makes them feel nice. That's not what this country is based on." @WattersWorld
3
30
161
There is a joy & creativity to the infinite computational canvas. It feels as interesting & fun as when I wrote my first BASIC games at age 8. It’s sad that younger geeks grew up in a time when the joy of coding has been captured by capitalistic, industrial labor economics.
7
7
169
9,020
1. Every quantization or boundary or framework we use to structure the world is imperfect, and misses fundamental aspects. 2. There is no past or future, only an ever-changing present. Time is an illusion, and is generally frequency, not duration. 3. Choice is not freedom.
What is the main insight that made you who you are today?
9
41
163
35,737
“Look at this here dumb, racist bitch, all full of birthright.” There, was that a sufficiently Americanized expression of contempt? I came here to the US from China, when it was a third world dump. My apt didn’t have potable water. Behold my successful assimilation.
Most third-world migrants can not assimilate into civil societies. Prove me wrong.
4
14
157
I hate how cynical this sounds, but the reality is quite undeniable: The point of the The News isn’t to inform people. It’s to make people *feel* informed. The product is an emotional one.
11
33
161
2020’s season finale is showing us the existence of epistemic closures so tight that even intelligent people will stick their heads into themselves in awkward ways, so as to turn into human Klein bottles of stupidity. Not a good look for Scott. (Then again, it never was.)
Democrats are employing some excellent brainwashing technique to defend the election as fair. Here are some of their tricks. 1. "Refuses to concede" is making you think past the sale that Trump's legal challenges will fail. This is their main persuasion trick. continued...
50
13
149
My tensors bring VCs to the yard And they’re like, “The math is too hard!” Damn right, the math is too hard I can teach them, but I’d have to charge
5
17
148
Replying to @chrisalbon
We’ll make an open source one
5
5
159
\o/ Great news, everybody! Pandas version 1.0 is here - Which I think means it is now ready for general use. 😉 Really happy @Anacondainc has been able to support some of the architectural enhancements. Congrats to the whole team!
We’re so excited for the milestone the pandas team has reached with the release of 1.0 today! bit.ly/36HmP1G #datascience @pandas_dev
1
37
153
Your daily reminder that static patterns result from flow and constraint Everything is simultaneously static and dynamic. Any perceived paradox is the result of sampling bias and compression.
4
39
150
It finally happened: Node.js has brought my child to tears. He was attempting to implement a Hangman game, and... Javascript punched him in his little geek soul. All the HackerNews bros that flamed me for creating @pyscript_dev: This is why. The tears of my children. ✊ 🐍
5
12
140
Looks like 2019 will be the year of “Linux on the Desktop”.exe
3
30
142
8 yr old: "What does DNA stand for?" Me: "Deoxyribonucleic acid" 8 yr old: "If my children ask me that, I'll just say 'Ask your grandfather', because I cannot say that."
13
143
It's June 4 (in China). 🇨🇳 🗽 Never forget.
5
24
137
If Trump fires Mueller, I'm going to explore options to take a (partial) sabbatical to get involved in political activism&apply my tech skills directly to advancing the cause of GOTV for midterms, exposing Russian propaganda, & destroying Facebook & Twitter. #Patriot #USA
8
22
143
Wow. Mad props to @Microsoft for contributing to OpenStreetMap and open data like this! blogs.bing.com/maps/2018-06/…
1
53
141
Meh. The surveillance infrastructure you can see is not the surveillance infrastructure you should worry about. It's quite telling that people are more worried about this robot than the phones in their pockets.
6
16
138