Sean Davidson

@silvertrade The radical right lunatics might not like it much either

"GOD MIGHT BE PLAYING HIS TRUMP CARD!" 🚨President Trump posts "God might be playing his TRUMP card!" with an image depicting Jesus Christ with His arm around Trump‼️ "The Radical Left Lunatics might not like this, but I think it is quite nice!!!"

@Bogachan_1971 This one will go down in history... but not for the reason Trump thinks

"Market" is trying to trade the posts and speeches of a lunatic who thinks he is Jesus Christ..... I guess Chinese and Russians still have doubts... they think there is a ploy here... maybe a bait... there is nothing. What You See Is What You Get....

@LukeGromen That's two things. Stop it - you're confusing him.

1. Then open the waterway. 2. Tell us what China and/or Russia have threatened in retaliation for flattening Iran.

@silvertrade Worst. Negotiations. Ever.

🚨PRESIDENT TRUMP WARNS US ATTACKS ON IRAN WILL BE "BIGGER, BETTER, & STRONGER THAN ANYONE HAS EVER SEEN BEFORE" IF TEHRAN VIOLATES THE CEASEFIRE⚠️ "IT WAS AGREED, A LONG TIME AGO, & DESPITE ALL OF THE FAKE RHETORIC TO THE CONTRARY - NO NUCLEAR WEAPONS AND, THE STRAIT OF HORMUZ WILL BE OPEN & SAFE." We suspect the IRGC is not actually interested in a ceasefire that meets ANY of Trump’s conditions, & we will be back to square one before the 2 weeks is up…

@DarioCpx "we are negotiating with everyone who is not dead yet and will most definitely find a solution, but if we don't we'll kill them just to make sure"

Taco man is on fire today

@Panoptica_ai @RickLakeNAlpha Thank you for this fascinating and inspiring article. I was reminded of children's stories, which come closer to the oral tradition than adult novels. Kipling's "great grey-green, greasy Limpopo River, all set about with fever trees" came especially to mind.

"The singer of tales is at once the tradition and an individual creator." "Singer of Code" by @RickLakeNAlpha. panoptica.ai/singer-of-code/

"Like the guslari, these AI systems don't retrieve pre-written text. They compose in performance: generating language token by token, selecting from learned patterns, creating outputs that are simultaneously novel and formulaic."

Wow! This is extraordinarily good. A novel and incredibly interesting insight into the creative mind of LLMs. If you have an interest in creativity, writing and LLMs, you must read it!

@Seamus_Malek @EpsilonTheory Good god

I don't think a more embarrassing and humiliating thing has ever been produced before by any government in human history. I somehow might be underselling it.

This is an epic post - about where the responsibility really lies with AIs (LLMs) and permitting them to undertake mass surveillance of (US) citizens and control fully autonomous lethal weapons. A very important discussion.

One week left to submit your abstract for the European ISHR 2026 meeting in Birmingham. Use the link below uobevents.eventsair.com/39th-ishr-es/a… @escardio @AHAScience @ISHR_ES #Cardiotwiter #phdone

Last chance to register for our #webinar "The Beating Edge of Human Centric Cardiovascular Models" - tomorrow Thursday, February 26th at 14:00 CET. Register with the QR code or here: us02web.zoom.us/webinar/regist… #cardiovascular #research #webinar

@Ross__Hendricks "potentially weeks long"

Launching a full scale war with Iran will go down as the single worst decision of the Trump administration And no one owns enough #oil

@izakaminska @TheMichaelEvery Insightful comment and the "ant mill" analogy is brilliant. I find the value of AIs responses is highly dependent on the question asked. They often don't volunteer what you're missing unless you explicitly ask for it (the "blindspot"), so you can get stuck in rabbit holes.

I think this is broadly rubbish. Have you read AI content? Don’t get me wrong, AI is super useful for processing large amounts of data. Has seriously revolutionised my ability to process transcripts. But I for one can immediately tell when someone has just plopped their ideas into AI and asked it to generate a text or an analysis. There is a weird synthetic tone to it. The language is devoid of character. It makes it difficult to read to the bottom. And I sometimes wonder if the author himself has even read everything they’re putting out. It is also untrustworthy. I appreciate it’s getting better and better and there are fewer hallucinations. But it is also possible that it has just become better at disguising its lies and fabrications. I wouldn’t be at all surprised if a number of major facts that were totally made up by AI have now been absorbed into the media complex as truth because of this effect. The ant mill risk is very real. And when it hits producers of source data, it will open the door to total disreality. We already can’t agree on major phenomena being real or fake. We already can’t agree if x thing is good or bad. If we all stop observing and judging the data on a distributed basis ourselves, and just accept that what AI says is true, we will abandon the formation of consensus reality. The AI’s reality, based increasingly on the biases of its original source code, will dictate everything. There will be little capacity for course correction if we allow AI to lose the human reference rate. I predict left to its own devices it will inevitably collapse in on itself or its output will become so absurd as to be rendered completely useless. Employers who lay off humans and replace them AI are the ones the stock market should be discriminating against. ESG style. Why? Because they are collapsing their own industry in the long run. They are exposing their sector to “ant mill” collapse risk. They are sending a signal that they are prepared to give up on human supervised quality control. Companies should care about this because their products will then no longer cater to the real desires and needs of people, but of AI’s perception of those needs. AI must serve humans, not humans AI, to avoid dystopia. And for human supervised quality control to be a thing, humans need to be trained in the basic tasks the system is outsourcing to AI. We all use calculators. But we also still teach kids mental arithmetic. From now on company value should be linked to how committed a company is to maintaining human supervision, on the basis that companies that go for short term profits over long term sustainability (by sacking humans and replacing them with AI) are eventually going to become peddlers of stranded assets. That should create a market for well trained and competent, but also TRUSTWORTHY people. People who don’t cut corners. Also, I predict, companies devoid of people won’t be able to compete with start ups. When AI lowers the barrier to entry for starting businesses, start ups will grow in highly productive but human symbiotic ways. I see this in my own business. AI has been very helpful at allowing me to do more with less. But the constraint on growth is now not enough me to supervise all the content I would like to put out. The only way I grow is by entrusting more high caliber human supervisors into my system who are aligned with my company mission, creative in complementary ways, and agree the mission adds value to human life. The only way AI can compete is by promising it will assert my values better in my system than other humans I entrust. But I don’t trust AI to be a better me. Not least because “me” is always changing as I adapt to new realities, other people and lessons learned. There is no guarantee any AI doppleganger of me will make the same choices. It cannot have the full spectrum of my experiences to feed off unless it decides to operate in an exact parallel reality.

We are seeing Wall Street pick off sectors it sees being hit by Claude. But what about Wall Street itself? For a sector that lives and breathes “Rate cuts!”, it argues for a lower cost of capital…. because there’ll be no research, sales, trading, deal-making, or back-office costs! 😵‍💫

Nice work just out in @JMCCPlus! Congratulations to the Authors... #CardioX #CardioTwitter @JMCCardiology

Stuck between a ROCK2 and a hard place? New research just out shows selective ROCK2 inhibition reduces microvascular obstruction after myocardial I/R but doesn’t limit infarct size. Highlights distinct roles for ROCK isoforms in #cardioprotection. 🫀🔬 doi.org/10.1016/j.jmcc…

@TheBMA @sugarydoc @drkevinfernando @RCPhysicians @WessexSpRs @Soc_Endo @BritRenalSoc @Elena_Armeni @dr_chrisallen @ProfBaldeweg @BMA_GP @youngdiabendo @RoySocMed @NephJC @RenalFellowNtwk @TheBJCA @rasprclub

@DarioCpx Congratulations on being proven right once again!

Silver just cross the historical 100$ mark and these are the Comex volumes so far today. Very few investors saw this coming, no retail mania in sight, just pure law of real demand and supply at play. FYI when Reddit squeezed silver in 2011, over 300,000 contracts traded daily

@KobeissiLetter nice overture to peace

BREAKING: President Trump threatens 200% tariffs on France after France's President Macron declines to join his "Board of Peace." Reporter: "Any response to Macron saying he will not join the Board of Peace." Trump: "If they feel hostile, I'll put a 200% tariff on his wines and champagnes."

Interested in obtaining a Cardiovascular Science MSc from UCL in London? Book a chat between 6 Jan to 24 April 2026, with Dr Petros Syrris, Director of the Cardiovascular Science MSc, to learn more about the programme, careers or studying at UCL. ucl.ac.uk/prospective-st…

Three £4,000 bursaries are available for students joining the MSc in Cardiovascular Science at UCL in September 2026. More details: lnkd.in/e5Ys9y-D #study #research #london