Homo-Sapiens
Dublin City, Ireland
Joined March 2009
- Tweets5.8K
- Following361
- Followers314
- Likes8.4K
And yet again IND royally loses #indvseng . Someone should be answerable. But the only answer we will get is historical stats.
86
SteamVR NEVER works when I start it after a couple of weeks. Each and every time there's a new kind of error on which I need to spend two hours or more.
29
Arjun Pakrashi retweeted
7,000 false positives per square millimeter. The culprit was the lab gloves.
University of Michigan researchers just upended a core assumption in microplastics science. Latex and nitrile gloves, worn by the scientists doing the measuring, shed stearate particles that look chemically identical to polyethylene. Standard infrared and Raman instruments can't tell them apart. The gloves were counting as plastic.
Seven glove types tested. All contaminated. The cheapest fix: switch to cleanroom gloves, which dropped false positives to around 100 per mm² vs. 7,000.
The "credit card per week" headline (5 grams, WWF/Newcastle 2019) has separate problems. A 2022 re-analysis found severe methodological errors in the original estimate. Actual measured intake is likely 100x lower.
None of this means microplastics are harmless. Last month's data on brain accumulation still stands. But the numbers driving the panic may have been measuring the scientists, not the environment.
Science catching its own errors is exactly how it's supposed to work.
Arjun Pakrashi retweeted
Hallucinated citations highest in social sciences preprints site - but, as I point out, part of the difference could reflect differences in moderation and publication stage, rather than something inherent to social sciences themselves.
More than 140,000 fake citations across four research repositories were identified in papers and preprints published in 2025 alone
go.nature.com/4uH7o54
Arjun Pakrashi retweeted
After seeing that Claude Mythos marketing turned out to be, as expected, a scam, I wanted to make a master list of tricks being used to market LLMs.
The master list includes statements directly from leadership in the companies or from the "organic marketing" of people on social media, along with an explanation on how the scam works. This is my first attempt, so likely incomplete.
The LLM Marketing Scams Master List v1:
"Two more weeks" - the models will be good enough someday soon to do what we claim.
"They're already good enough" - the models are already good enough to replace workers, but it hasn't happened yet because of x y z reasons.
"We just built God in the backroom, and no, you can't see it" - the models they built in private are actually capable of doing the things we have been waiting for, but they can't let us see them yet for x y z reasons.
"Actually they already have replaced jobs" - the layoffs that tech companies have been doing, citing AI as the reason, have already been replaced with current LLM tech, ignoring market conditions and past data on layoffs during such conditions.
"You just don't know how to use then as well as me" - the models are good enough, but esoteric prompt engineering is required to get these results, and no, I won't teach you.
"I built an app making big money with LLMs" - they claim they already have made startup companies, almost always SaaS companies, that are making them tons of money, but when you ask to see them, they won't show you.
"You aren't using the right model" - claims that you must be using the wrong model and need to use Open Claude 420b-parameter Gemini Plus Pro 6.9 with 4RealThisTime HomerSimpson agent mode enabled. Note that this will be used to attack every study on the effectiveness of LLMs, since studies take time to complete and publish, with new models releasing more frequently than it's possible to complete and publish a study
"You're falling behind" - claims that you need to use the bots now, even though they aren't good enough to fully automate any jobs, because otherwise, when the bots are good enough, you will lose your natural English skills required to prompt effectively.
"All these companies are using LLMs, so do you think you know better than they do?" - pointing to claims of large companies deeply invested in LLMs being a success saying that LLMs are being used effectively, with no viewable results in the speed and/or quality of their company's output.
"The benchmark score went up" - claiming improvements on the benchmarking tests given to their latest model, despite the training being specifically tuned to improve on these tests, and then conflating better benchmark scores with actually being more able to automate jobs or drastically improve worker productivity.
"It can now count the letters in Strawberry/can now do things it famously couldn't do previously" - saying that it can now count the letters in Strawberry or instruct you on how to use a cup without a bottom, etc. is often done to suggest increased reasoning for the LLM, but often involves just hard coding an answer into the service.
"It has escaped our control" - saying that they cannot control the LLM, implying it is conscious or living to some degree when really it just said words that it wasn't supposed to or an agent used an app that wasn't intended by the user's prompt when next-token predicting
"It's feeling sad/scared/happy/angry, suggesting it is conscious" - they ask the LLM how it is feeling, and it next-token predicts a response that includes an emotion felt by humans, since training data is from human conversations online.
"Costs are going down/the LLM service is profitable" - ignores training costs and capex for hardware, usually just referring to inference being profitable, which isn't even true in many cases. Training and capex is 95%+ of the total costs to serve the models.
Did I miss any?
ENG vs IND. Extreme game, never seen such a game where each and every ball mattered. It's a shame that Jacob Bethell was not the man of the match. The undefeated (run-out doesn't count) run machine.
54
Arjun Pakrashi retweeted
Delighted to present @UCDCompSci research at IEEE CISP-BMEI 2025 in Qingdao with @phoxis. Our works span AI for environmental forecasting, spectral translation, glacier reconstruction & efficient video recognition, advancing @ucddublin's vision for sustainable digital innovation.
Arjun Pakrashi retweeted
Some people today are discouraging others from learning programming on the grounds AI will automate it. This advice will be seen as some of the worst career advice ever given. I disagree with the Turing Award and Nobel prize winner who wrote, “It is far more likely that the programming occupation will become extinct [...] than that it will become all-powerful. More and more, computers will program themselves.” Statements discouraging people from learning to code are harmful!
In the 1960s, when programming moved from punchcards (where a programmer had to laboriously make holes in physical cards to write code character by character) to keyboards with terminals, programming became easier. And that made it a better time than before to begin programming. Yet it was in this era that Nobel laureate Herb Simon wrote the words quoted in the first paragraph. Today’s arguments not to learn to code continue to echo his comment.
As coding becomes easier, more people should code, not fewer!
Over the past few decades, as programming has moved from assembly language to higher-level languages like C, from desktop to cloud, from raw text editors to IDEs to AI assisted coding where sometimes one barely even looks at the generated code (which some coders recently started to call vibe coding), it is getting easier with each step.
I wrote previously that I see tech-savvy people coordinating AI tools to move toward being 10x professionals — individuals who have 10 times the impact of the average person in their field. I am increasingly convinced that the best way for many people to accomplish this is not to be just consumers of AI applications, but to learn enough coding to use AI-assisted coding tools effectively.
One question I’m asked most often is what someone should do who is worried about job displacement by AI. My answer is: Learn about AI and take control of it, because one of the most important skills in the future will be the ability to tell a computer exactly what you want, so it can do that for you. Coding (or getting AI to code for you) is a great way to do that.
When I was working on the course Generative AI for Everyone and needed to generate AI artwork for the background images, I worked with a collaborator who had studied art history and knew the language of art. He prompted Midjourney with terminology based on the historical style, palette, artist inspiration and so on — using the language of art — to get the result he wanted. I didn’t know this language, and my paltry attempts at prompting could not deliver as effective a result.
Similarly, scientists, analysts, marketers, recruiters, and people of a wide range of professions who understand the language of software through their knowledge of coding can tell an LLM or an AI-enabled IDE what they want much more precisely, and get much better results. As these tools are continuing to make coding easier, this is the best time yet to learn to code, to learn the language of software, and learn to make computers do exactly what you want them to do.
[Original text: deeplearning.ai/the-batch/is… ]
Australia is demonstrating why they are the best team (statistically) again. #AUSvsENG #ChampionsTrophy2025
110
Arjun Pakrashi retweeted
Yesterday, I got a DM on Twitter from someone asking for help. His brother was being denied an admit card for his board exam because their school hadn’t received the pending ₹54,000 fee. When I asked for more details, he told me that his parents were at the school, trying to convince them to issue the admit card since the exam was just three days away. But the school refused.
I couldn’t just sit back, so I decided to help. @MamoniOjha and @sny_259 joined in too, and we passed the message to @arunbothra Sir, the ADGP of that respective place. He asked for all the details and started verifying everything. Meanwhile, something incredible happened.
An anonymous Twitter account reached out, asked for the admission number, and without saying much, directly paid the fee on the school’s website. No questions, no expectations, just a simple "All the best!" for the student.
With all the negativity on Twitter, moments like these that show there are still kind people out there, strangers who step up to help without knowing each other.
Sourav’s brother finally got his admit card and will be able to write his future. Sourav is also preparing for JEE and scored 97 percentile in the January attempt.
This act of anonymous might just help his family move one step forward to achieve their dreams.
Arjun Pakrashi retweeted
🛑Monte Carlo for Prediction Intervals? Think Again. 🛑
Here’s why:
Using prediction errors from a model trained on training data is a recipe for overfitting and guarantees incorrect intervals.
In 2024, with the power of complex models, you can easily achieve near-zero errors by overfitting almost any dataset.
But what happens when you sample these artificially minimized errors using Monte Carlo to predict intervals for unseen data?
What could possibly go wrong?
Spoiler: Everything.
Prediction intervals derived this way are not just unreliable—they’re misleading.
Overfitting ensures that these intervals bear little resemblance to real-world uncertainty, leading to false confidence and poor decision-making.
If you're building prediction intervals, always start with calibration on unseen data and use robust methods that account for the true distribution of errors—not ones distorted by training data overfitting.
Let’s stop misapplying tools and start building trust in ML predictions the right way.
#MachineLearning
#MonteCarlo
#PredictionIntervals
Arjun Pakrashi retweeted
Had a great time presenting our research at the 17th IEEE conference CISP-BMEI 2024 in Shanghai, China! Arjun Pakrashi (@phoxis) and I shared insights on solar forecasting, spatial prediction models, and CO₂ emissions in Ireland. Great to engage and exchange ideas!#remotesensing
ALT (From L to R): Arjun Pakrashi and Soumyabrata Dev.
Arjun Pakrashi retweeted
Ola electric hasn’t disclosed any plan to issue refunds or put an end date to current customer complaints. We don’t even know
If there is a plan…
All I can do is Let @bhash know that he has to put out a public plan which doesn’t include employing me.
I didn't know the issues about Ola's EVs. It's so severe that any other automotive companies or technology companies would have recalled their products to protect their customers. But it looks like Ola Electric is still just a marketing company who happens to sell EVs. @bhash
1
122
Arjun Pakrashi retweeted
Gino's case against us has been dismissed.
Scientists cannot effectively sue other scientists for exposing fraud/errors in their work.
Those who work to correct the scientific record can sleep better tonight. Those who don’t want it corrected, well, I don’t care how they sleep.