22

Joined August 2017
Vindiw Wijesooriya retweeted
Gemini 3 Deep Think generated a real-time 3D WiFi radar that maps every network around you as glowing nodes in a Matrix-style space — in one shot. It used Pearson correlation to infer which APs are physically close, since RSSI alone isn't enough.
178
831
77
10,223
693,790
Vindiw Wijesooriya retweeted
Starlink now available in Sri Lanka! 🇱🇰
Starlink's high-speed, low-latency internet is now available in Sri Lanka! 🛰️🇱🇰❤️ → starlink.com/srilanka
10,451
11,532
1,140
150,196
43,507,350
Vindiw Wijesooriya retweeted
"I predict that by 2042 95% of humanity's energy usage will be downstream of solar." — Casey Handmer @CJHandmer
10
45
15
340
65,127
Vindiw Wijesooriya retweeted
+1 for "context engineering" over "prompt engineering". People associate prompts with short task descriptions you'd give an LLM in your day-to-day use. When in every industrial-strength LLM app, context engineering is the delicate art and science of filling the context window with just the right information for the next step. Science because doing this right involves task descriptions and explanations, few shot examples, RAG, related (possibly multimodal) data, tools, state and history, compacting... Too little or of the wrong form and the LLM doesn't have the right context for optimal performance. Too much or too irrelevant and the LLM costs might go up and performance might come down. Doing this well is highly non-trivial. And art because of the guiding intuition around LLM psychology of people spirits. On top of context engineering itself, an LLM app has to: - break up problems just right into control flows - pack the context windows just right - dispatch calls to LLMs of the right kind and capability - handle generation-verification UIUX flows - a lot more - guardrails, security, evals, parallelism, prefetching, ... So context engineering is just one small piece of an emerging thick layer of non-trivial software that coordinates individual LLM calls (and a lot more) into full LLM apps. The term "ChatGPT wrapper" is tired and really, really wrong.
I really like the term “context engineering” over prompt engineering. It describes the core skill better: the art of providing all the context for the task to be plausibly solvable by the LLM.
531
2,052
614
14,330
2,402,589
Vindiw Wijesooriya retweeted
Something I find obvious but am realizing is very non-consensus: almost no one should ever go to college, and people should start working at age 14-16 instead (what I did) 1/ it's a very modern idea that you're not supposed to start your active life until you're 23-25, and it's quite harmful imo. You have all these 20-somethings that are basically babies — and worse: not only do they not realize they're late, they actually are lulled into thinking that they have all the time in the world ahead of them, that "their 20s are for fucking around." This is also an obvious and big contributor to the fertility crisis. 2/ People think that this is justified by modern jobs being much more complex / intellectually demanding. But Ben Franklin started working full-time at 12, Carnegie at 13, Rockefeller at 16. You can't exactly argue these folks weren't smart or sophisticated, or their jobs not extremely complex. 3/ More pragmatically, ask yourself who you'd rather hire between a 22 year-old who's been working for 7 years, and one who's straight out of college. I view the former as basically senior, probably even ready to lead a team — the latter as next to useless (which is why new grads struggle to find a job so much). It's basically common knowledge that college years are mostly spent fucking around / going to parties — insane that we accept this as normal, and that's before even talking about the hundreds of thousands of dollars in debt that people get into for these party years. So, not only people spend less of their time acquiring useful skills while in college — what little they spent is _spent less well_, because it's A) removed from a concrete context B) taught by far-left boomers with very outdated skills (recently met a new grad from a design school who was using Photoshop(!!) to design mobile apps).
51
51
15
764
72,985
Vindiw Wijesooriya retweeted
I've been on this little blue dot for 45 years now. Seen computers go from 1 Mhz C64s to M chips. Seen the internet rise in the 90s. Mobile in the 00s. And this is the first time I've felt I have no clue what the world will look like in four years. Exhilarating time to be alive.
202
447
81
7,002
443,649
Vindiw Wijesooriya retweeted
X on a silly week is so much better than X on a tribal week
51
63
8
2,471
191,355
looks great as a wallpaper!
Replying to @OwenSparks
ghiblified versions of these iconic images
1
3
362
Vindiw Wijesooriya retweeted
Replying to @GrantSlatton
it's so good!
1
16
1,602
Vindiw Wijesooriya retweeted
man cloudflare is so insanely good
121
38
12
1,658
147,573
Vindiw Wijesooriya retweeted
most of the world starts making sense when you realize that modern leftists don’t hate oppression, they hate excellence. nothing enrages them more than competence because it exposes their own mediocrity. they aren’t driven by compassion, only resentment for what they can never be. their entire ideology is a war on hierarchy, not because hierarchy is unjust, but because they know they’ll always be at the bottom unless they can manipulate the system. a simple but great world model: the good is whatever leftists oppose
It's so creepy that Elon Musk owns most of the satellites in space. No one should be allowed to own most of the satellites in space.
229
1,689
167
15,861
810,215
first impressions: i love it
We're opening limited access to a research preview of a new agentic coding tool we're building: Claude Code. You'll get Claude-powered code assistance, file operations, and task execution directly from your terminal. Here’s what it can do:
1
222
Vindiw Wijesooriya retweeted
Agency > Intelligence I had this intuitively wrong for decades, I think due to a pervasive cultural veneration of intelligence, various entertainment/media, obsession with IQ etc. Agency is significantly more powerful and significantly more scarce. Are you hiring for agency? Are we educating for agency? Are you acting as if you had 10X agency? Grok explanation is ~close: “Agency, as a personality trait, refers to an individual's capacity to take initiative, make decisions, and exert control over their actions and environment. It’s about being proactive rather than reactive—someone with high agency doesn’t just let life happen to them; they shape it. Think of it as a blend of self-efficacy, determination, and a sense of ownership over one’s path. People with strong agency tend to set goals and pursue them with confidence, even in the face of obstacles. They’re the type to say, “I’ll figure it out,” and then actually do it. On the flip side, someone low in agency might feel more like a passenger in their own life, waiting for external forces—like luck, other people, or circumstances—to dictate what happens next. It’s not quite the same as assertiveness or ambition, though it can overlap. Agency is quieter, more internal—it’s the belief that you *can* act, paired with the will to follow through. Psychologists often tie it to concepts like locus of control: high-agency folks lean toward an internal locus, feeling they steer their fate, while low-agency folks might lean external, seeing life as something that happens *to* them.”
Intelligence is on tap now so agency is even more important
1,949
9,324
1,874
49,996
11,479,429
Vindiw Wijesooriya retweeted
I still cannot believe that I can: - look at a world map and tap anywhere to zoom in at street level - instantly access any song, book, movie, tv show, or podcast ever made - have any conceivable question and get an immediate answer or video explanation - take a photo or video wherever I am and add it to my massive, searchable, always accessible personal archive - video call anyone in my life, at anytime, no matter where they are - watch live sports on a little wireless glass rectangle - type out these thoughts and have them read by thousands of people, all over the world, a few seconds later
1,292
5,133
606
52,863
16,241,199
Vindiw Wijesooriya retweeted
New 3h31m video on YouTube: "Deep Dive into LLMs like ChatGPT" This is a general audience deep dive into the Large Language Model (LLM) AI technology that powers ChatGPT and related products. It is covers the full training stack of how the models are developed, along with mental models of how to think about their "psychology", and how to get the best use them in practical applications. We cover all the major stages: 1. pretraining: data, tokenization, Transformer neural network I/O and internals, inference, GPT-2 training example, Llama 3.1 base inference examples 2. supervised finetuning: conversations data, "LLM Psychology": hallucinations, tool use, knowledge/working memory, knowledge of self, models need tokens to think, spelling, jagged intelligence 3. reinforcement learning: practice makes perfect, DeepSeek-R1, AlphaGo, RLHF. I designed this video for the "general audience" track of my videos, which I believe are accessible to most people, even without technical background. It should give you an intuitive understanding of the full training pipeline of LLMs like ChatGPT, with many examples along the way, and maybe some ways of thinking around current capabilities, where we are, and what's coming. (Also, I have one "Intro to LLMs" video already from ~year ago, but that is just a re-recording of a random talk, so I wanted to loop around and do a lot more comprehensive version of this topic. They can still be combined, as the talk goes a lot deeper into other topics, e.g. LLM OS and LLM Security) Hope it's fun & useful! youtube.com/watch?v=7xTGNNLP…
753
2,888
605
20,019
2,396,041
Vindiw Wijesooriya retweeted
Humanoid robots and drones will both be absolutely everywhere 10-20 years from now. Delivery drones flying overhead, robot janitors, baristas, grocery store employees guiding you to the right aisle, and (eventually) housekeepers—will all seem as normal as smartphones do today.
1,581
1,028
512
9,289
11,003,602
Vindiw Wijesooriya retweeted
the reason people cannot fathom young people taking leadership roles or responsibility is because it threatens their very existence. if they can't rely on age or "muh years of experience" for their positions, they wouldn't have it because they have no skills
198
609
63
8,676
285,152
Vindiw Wijesooriya retweeted
you can learn a lot about a person by how they refer to Elon
44
61
10
1,506
78,287
Vindiw Wijesooriya retweeted
There's a new kind of coding I call "vibe coding", where you fully give in to the vibes, embrace exponentials, and forget that the code even exists. It's possible because the LLMs (e.g. Cursor Composer w Sonnet) are getting too good. Also I just talk to Composer with SuperWhisper so I barely even touch the keyboard. I ask for the dumbest things like "decrease the padding on the sidebar by half" because I'm too lazy to find it. I "Accept All" always, I don't read the diffs anymore. When I get error messages I just copy paste them in with no comment, usually that fixes it. The code grows beyond my usual comprehension, I'd have to really read through it for a while. Sometimes the LLMs can't fix a bug so I just work around it or ask for random changes until it goes away. It's not too bad for throwaway weekend projects, but still quite amusing. I'm building a project or webapp, but it's not really coding - I just see stuff, say stuff, run stuff, and copy paste stuff, and it mostly works.
1,475
3,633
2,549
34,158
7,371,660