Romjan Ali

1.7K posts

Romjan Ali banner
Romjan Ali

Romjan Ali

@romjanali

Life science commercialisation. Care about using healthcare innovation to improve outcomes for underserved populations.

London, England Katılım Eylül 2009
106 Takip Edilen195 Takipçiler
Romjan Ali retweetledi
Shivani Misra
Shivani Misra@ShivaniM_KC·
Crazy obesity guidelines for Mounjaro on the NHS, in which a patient can have BMI> 40kg/m2 OSA Dyslipidaemia Hypertension and Pre-diabetes Started Tirzepatide privately & lost 15kg but unaffordable so got referred to Tier 3, failed eligibility for Mounjaro (no T2D) and got switched to Wegovy starting at 0.25mg!!! And regained all the weight lost What are we doing here???
English
38
23
301
53.5K
Romjan Ali retweetledi
Dr. Michael B. Riordan
Dr. Michael B. Riordan@michael_riordan·
The night before the story about @Apple introducing ID checks broke, I upgraded my phone to find myself locked behind an age verification system. Despite the fact that I am 38, there was no obvious way to verify my age. The system didn’t explain the consequences of this, so I was genuinely worried that I wouldn’t be able to access my phone. I was not unique —- many people have reported being locked out of their phones: this is not strictly true, but it is what the messaging very strongly implies. Given how much we rely on our phones, the response is similar to those you might expect if you were locked out of your house: yes, you can hire a locksmith, but what if they can’t get there for hours - or turn out to be a con artist? It was only a day later that this was reported in the media, and it was two days later that Apple published an article clarifying what was blocked and the forms of ID required. I have still not been able to verify my age — of the forms of ID, the only real option available to me is a CirizenCard run by @getyoti but the application costs money and takes weeks to complete. I need to trust that @getyoti are legit and will not refuse — given that I have criticised the company in blog posts this is not a certainty. Not being able to prove that I am who I say makes me anxious. As a result my phone literally thinks I am a child: currently the restrictions are limited, but they are demeaning: if someone sends me a nude photo for example (not something that has happened to me in many years), a message would pop up asking me if I want emotional support or need to contact an ‘adult’. It looks likely that when the UK government implements restrictions on social media within the next year or so, I will be blocked from much of the Internet if @getyoti decide not to verify me, as they are under no legal obligation to do. What makes this all the worse for me is Apple’s response. At the outset, I wrote to @tim_cook setting out all of my concerns, including (briefly) the fact that I was visually impaired, which tends to make some of these ID checks problematic. I told Tim I would share any response if I didn’t receive a response within 7 days. Yesterday (5 days) I got what looked like automated response from Apple ‘accessibility’ team which gave me a case number, but ignored all the points in the email, and made no suggestion of how I might proceed. The deliberate gaslighting of disabled users does seem to undermine irrevocably the company’s case their products are accessible and I’d urge other disabled users to think twice before giving this company their business.
English
13
48
190
20.2K
Romjan Ali retweetledi
Saloni
Saloni@salonium·
This isn't what the trial showed. The control group was also recommended to increase fluid intake, and did; the intervention group had much more intense guidance (e.g. financial incentives and regular text messaging), and drank even more. So we can't conclude that hydration is ineffective; the benefit may have already been saturated.
Eric Topol@EricTopol

If you've had a kidney stone, you've been advised that the most important thing to prevent another bout is to increase hydration. Now a randomized trial of hydration in over 1600 participants showed no benefit, despite evidence of increase during volume. thelancet.com/journals/lance…

English
5
15
215
53.7K
Romjan Ali retweetledi
Sam Bowman
Sam Bowman@s8mb·
The Tory promise to scrap the higher student loan interest rates for higher rate earners really shows a problem they have – they introduced these higher rates! They were part of a big swathe of punitive measures for higher earners: - Higher student loan interest rates above £50k/year - Child benefit withdrawal above £50k/year - Childcare subsidies withdrawn at £100k/year - IR35 on contractors - Pension allowance cut above £200k - Higher stamp duty on second homes - The pension Lifetime Allowance cut to £1m (abolished by Truss, thankfully) - Dividend allowance cut from £5k to £500 - CGT allowance cut from £12,300 to £3k - Buy-to-let mortgage interest deductions capped at 20% Plus fiscal drag, aka letting inflation push more people into higher tax bands. If the tax thresholds had been raised in line with inflation between 2010 and 2024, they would be: - Higher rate (40p rate) threshold: £50k → £66k - £100k threshold: £100k → £150k - Additional rate (45p rate) threshold: £150k → £225k These measures were probably viewed as a relatively easy way to raise revenues, plus a way to mitigate the allegation that the Tories were the party of the rich. But apart from creating a *much* more progressive system than it looks like we have, making the country less hospitable for highly productive workers, they have probably helped to turn these voters against the party. Higher earners are now no more likely to vote Tory than anyone else. And now they need those voters again, all they can do is promise to undo some of the things *they did*. The risk of taking part of your base for granted!
Sam Bowman tweet mediaSam Bowman tweet media
English
28
63
344
39K
Romjan Ali retweetledi
Saloni
Saloni@salonium·
The cost of sequencing a human genome has fallen over 100,000 fold in nominal terms since 2001. In a new visualization, I've added some of the key advances in sequencing during that timeline:
Saloni tweet media
English
23
164
742
41.8K
Romjan Ali retweetledi
Nabeel S. Qureshi
Nabeel S. Qureshi@nabeelqu·
Very interesting piece in which an author fine-tunes a model on her writing, asks her friends to guess which is her and which the AI, and nobody guesses correctly. This doesn't make me feel particularly doomer-ish on art/writing, though. Rough thoughts on why: - The generated texts are all pastiches of human authors. The study shows experts prefer AI-generated pastiches to human-written ones. - The concern expressed in the piece is: what if an author just trained an AI on their own corpus, then used that fine-tuned AI to generate a ton more stories in their style, under their name? - But I think this concern misses part of what we value in art, which is not just 'good execution' but also novelty. If you generated more finetuned early Kanye you'd get more "Graduation" but you wouldn't get "Yeezus". If you did more early Joyce you'd get more "Dubliners" but no "Ulysses". If you did more early Dylan you'd get more acoustic folk songs but none of the electric stuff. etc. Part of being an artist is to make progress in art. - I also think we inherently value human involvement in the creation of art, just like we prefer watching human chess to AI chess (even if the humans are playing out ideas they got from AIs, which is almost always the case nowadays!). If nothing else, it's a filtering mechanism on your attention for when generating floods of competently-executed short stories becomes possible. - So I expect that art, at least at the high end, will remain 'centaur-like' for a long time, with AI as collaborator/executor and humans continuing to be the main director/producers.
The New Yorker@NewYorker

Is it possible to train A.I. models to produce great literature? In a recent experiment, M.F.A. students preferred A.I.-generated passages over passages written by their peers. newyorkermag.visitlink.me/Qgd6Ab

English
13
20
323
42.4K
Romjan Ali retweetledi
Saloni
Saloni@salonium·
Novels' sentences are much shorter than they used to be. Part of this is about punctuation: we use periods as people used to use semi-colons. And shorter sentences aren't necessarily simpler to read, since jargon can condense complicated ideas. But plain language writing has also become more common. More people write as they speak. That's why English prose has become easier to read, argues @HenryEOliver, in a great new piece for Works in Progress. worksinprogress.co/issue/the-logi…
Saloni tweet media
English
61
211
1.6K
292.3K
Romjan Ali retweetledi
Ananyo Bhattacharya
Ananyo Bhattacharya@Ananyo·
This is getting silly… Wanna buy the intellectual biography of one of the most consequential figures of the 20th century for less than $4!?
Ananyo Bhattacharya tweet media
Ananyo Bhattacharya@Ananyo

Folks! Amazon.com has done it again! Kindle edition of 'The Man from the Future' on sale for less than $5! That's 75% off. Want to find out what the fuss over John von Neumann is really about and what he actually did? You can now for $4.80! #SalesRank" target="_blank" rel="nofollow noopener">amazon.com/dp/B098TYZN67#…

English
1
2
7
1.5K
Romjan Ali retweetledi
Garry Tan
Garry Tan@garrytan·
If you *only* used SAT to admit to elite colleges, share of admits from top 1% income falls 15.8% → 9.9% and representation from <$200k rises by +8.8%, with no reduction in post-college outcomes. It's 'holistic review' and 'ban SAT' policy that allows the most wealthy and powerful to virtue signal while getting an edge.
Garry Tan tweet media
English
405
2.7K
18.4K
3.1M
Romjan Ali retweetledi
Dr. Michael B. Riordan
Dr. Michael B. Riordan@michael_riordan·
First time in many years been refused entry from a venue in @CityWestminster because bouncers mistook my disability for inebriation —and that I suspect is just because I go out less! I know jasmine tea can be strong, but…. Won’t embarrass the venue — heard from my friend that there were only 5 others inside — but I’d be interested to hear from the regulator @SIAuk what are rules for training bouncers around disability awareness and what are the penalties for getting it wrong? #DisabledNotDrunk
English
0
2
4
595
Romjan Ali retweetledi
Dr. Michael B. Riordan
Dr. Michael B. Riordan@michael_riordan·
The Online Safety Bill was conceived by the last government, and its failures were the result of decisions made by Tory ministers, but Labour MPs wholeheartedly endorsed it, and Kyle has staked his reputation by it. It is bad by design because it confuses three things — the admirable goal of preventing kids from accessing content which existing laws mean it is illegal for them to access, the goal of limiting their access to content which was not previously deemed illegal, but which the government now deems harmful; and the totally unjustified goal of making it harder for adults to find content which is not illegal but which the government now apparently deems immoral. And it is bad by design because it ignores the evidence that blocking categories of user from categories of content just leads to an uptick in use of parts of the web that are freer of regulation. But rather than take the blame for a #BadLaw that has rightly been criticized on both left and right s for curbing our free access to information, and contorting our democratic norms, @peterkyle seems intent on channeling the spirit of Lord Longford. He first accused his critics of supporting pedophiles, and now wants to shift the blame to others who are just trying to implement his laws
Dr. Michael B. Riordan tweet mediaDr. Michael B. Riordan tweet media
English
3
6
10
1.1K
Romjan Ali retweetledi
William Dalrymple
William Dalrymple@DalrympleWill·
"Gazans 'wasting away' as mass starvation spreads, more than 100 humanitarian groups warn" (And still @Keir_Starmer hasnt said a word about this, the biggest moral failure of the West in our time) bbc.com/news/articles/…
English
75
612
1.4K
52K
Romjan Ali retweetledi
Dr Phil Hammond 💙
Dr Phil Hammond 💙@drphilhammond·
In today’s @PrivateEyeNews, I reveal why Dr Svilena Dimitrova (@NeoDoc11), a consultant neonatologist and instructed expert for Letby, has referred the case of Baby O back the Cheshire police and asked them to investigate the possibility of gross negligence manslaughter.   Her detailed report, prepared with another neonatologist, was discussed at a press conference in December, which she did not attend as she was working on other reports. It focused on injury to the liver. However, a bigger error was that the ventilation pressures were set so high, the heart couldn’t function properly. In Dr Dimitova’s view, this would have led to the death of the baby irrespective of what happened to the liver. 

A chest x-ray clearly shows hyperinflated lungs and the liver pushed down into the region where one of the consultants stuck a needle and drew back blood. A haemoglobin measurement shortly afterwards shows a catastrophic halving, indicating a major bleed most likely from the needle breaching the capsule of the liver which was already under pressure because of a subcapsular haematoma. The significance of the haemoglobin drop before resuscitation was discontinued and the failure to correct the profound acute anaemia was not adequately examined by the prosecution experts, and neither were the high ventilation pressures even though they are clearly documented in the notes. 

Dr Dimitrova believes that if the ventilation pressures and loss of blood had been spotted and corrected, as they should have been,  the baby would have lived. She could find no evidence of deliberate harm to the liver and no evidence of venous air embolism. She has also reported her findings to the Cheshire coroner asking for an inquest into the death of Baby O, and to the GMC to assess any competency issues amongst the medical staff. 

Dr. Brearey and the Countess of Chester Hospital have declined to comment, and I am awaiting a comment back from prosecution pathologist Dr Marnerides, to see if this new information changes his views on the cause of death.
English
33
108
306
38.1K