Carlos del Ojo Elías

I'm sorry for people that actually rely on LLM agents' output without verification. I'll never trust code written by LLM agents unless I know what I am doing and have the documentation for what I want to achieve. This is how I use LLMs these days, of course my favourite Anthropic Claude: ✅ First, I set a super clear outcome of what I want to achieve. This way you can't sway my course from what I wanted. ✅ I get myself ready, open the documentation, and tell it to strictly follow the official guides. This way, I have been able to get the best out of Claude without hallucination. Solely allowing LLM agents to take total control is a recipe for later nightmares. You see below my example when I was trying to implement an HA VPN connection between AWS and Google Cloud. 🚀🚀 LLMs are a superb productivity boost, but how you use them makes all the difference. #softwareengineering #cloudengineering #coding #innovation #technology #llm #artificialintelligence

52

11 Comments

The simplest way to get your study documents organised is with an eTMF vault. There is just so much to remember to stay on top of your game with, potentially, hundreds of docs per trial - PDFs, spreadsheets, contracts, agreements, SOPs... the list goes on. Typical things I hear in this space are: 👉🏼 What documents do I need and when? 👉🏼 Who needs to be aware, access and sign certain docs? 👉🏼 How do I keep on top of version numbers? 👉🏼 I'm new to this - I can't keep up with all the acronyms! 👉🏼 What if I miss something? If you're adhering to best practices like ICH-GCP you'll want to do the right thing from the outset to avoid many headaches. Learn more here: https://buff.ly/4cyX321 #clinicaltrials #nutritionaltrials #humantrials #epro #ecrf #econsent #etmf #cognitivetests #wearables #randomization #ctms #decentralisedclinicaltrials #trialflare

2

Reminds of an old blog, thousand dollar genome but million dollar analysis. While the sequencing price continues to drop, analysis remains a giant mess, costing tens of millions of dollars to find any semblance of an insight! https://lnkd.in/g_w7vV3B

31

Had a great time joining Phil and colleagues 🎙️ podcasting like I used to do years ago, and sharing some highlights from the Nextflow Summit. Check it out as we reflect on the best moments from a great event! 🚀✨ #NextflowSummit #Podcasting #Highlights

10

How much do you care about your data going missing? Not much? How about your DNA? Atlas Biomed, a DNA testing firm, has vanished, leaving users unable to access their data. And according to the BBC, they have given their customers no warning and no update. There aren't any signs of the data being misused yet, but with such valuable data on the line, would you trust companies with your genetics if it is this easy to just disappear? As it becomes easier to give away data to companies, it becomes harder to trust them. Read more here in the BBC article: https://lnkd.in/eNk5Uiey

5

An interesting video of some 50 minutes from Gresham College in the UK about evolution and how randomness affects species evolution. "As you almost certainly learned early on in your scientific education, species survive and perpetuate in large part due to their fitness for their environment. However, there’s more to evolution than just natural selection. Robin May is professor of infectious disease at the University of Birmingham and professor of physic at Gresham College in London, and in this lecture at Gresham College from November 2023, he explains why small random events compounded over time (as in the case of genetic drift) or anomalous events that occur in an instant (as in the massive meteor strike that killed off the dinosaurs) should also very much inform how we understand evolution. Using a combination of easy-to-digest explainers and fascinating examples from Earth’s evolutionary history, May lays out the multitude of ways we should – and should not – understand the role random chance plays in the evolutionary process." #biology #technology https://lnkd.in/ejK5pEyZ

1

Here's a story about how I tried to find details of a clinical trial for a friend. This is what happened. 1. Found some trials on Cancer research UK website, no contact details 2. Located them on CT.gov - no site contact details, one had a generic email 3. Response: this is only for US patients (it's an extensive global trial) 4. Found a link on CT.gov to a publication 5. Traced an author on the publication through Pubmed and found an email 6. They referred me down the line to 3 other people 7. Found a contact for the other trial on an online ethics application 8. Finally spoke to study coordinator / PI. I work in this space and this was far too hard and took too long. #clinicalresearch #clinicaltrials #clinicalsites

19

1 Comment

Interesting article about time! I always appreciate fundamental physics and how it affects our lives. Einstein’s theory of relativity is essential for modern computers and the internet. Without it clocks on satellites wouldn’t be synchronised correctly with ground ones. Hence, gps and so many more technologies wouldn’t work. This article suggests that time is an illusion. Interesting new perspective. Let’s see if this theory will be proven to be real and more interestingly how it will affect our lives!

6

The opportunities of AI for scientific research are highlighted throughout this The Royal Society report and explored in depth through three case studies on its application for climate science, material science, and rare disease diagnosis. Alongside these opportunities, there are various challenges arising from the increased adoption of AI. These include reproducibility (in which other researchers cannot replicate experiments conducted using AI tools); inter-disciplinarity (where limited collaboration between AI and non-AI disciplines can lead to a less rigorous uptake of AI across domains); and environmental costs (due to high energy consumption being required to operate large compute infrastructure). There are also growing barriers to the effective adoption of open science principles due to the black-box nature of AI systems and the limited transparency of commercial models that power AI-based research. Furthermore, the changing incentives across the scientific ecosystem may be increasing pressure on researchers to incorporate advanced AI techniques at the neglect of more conventional methodologies, or to be ‘good at AI’ rather than ‘good at science’. These challenges, and potential solutions, are detailed throughout this report in the chapters on research integrity; skills and interdisciplinarity; innovation and the private sector; and research ethics. 

24

Help support the amazing work being done by a very good friend of mine as he continues to innovate within the blossoming field of breathwork science 🔬 If you have benefited from a mindfulness, yoga, or meditation practice, then you already know just what kind of impact breathing can have on one’s wellbeing🧘‍♀️🧘‍♂️🧘 Guy and colleagues in his field are not only helping us to understand the science behind why breathing exercises can have such powerful positive impacts on the mind and body, but are trying to go beyond, striving to leverage said science so we can more easily attain and extend the benefits of breathwork 💪 Have a look, consider contributing, and sponsor the next major advance in this growing field 🌬️

7

Rory Cellan-Jones kickstarts his AI in the NHS series. The focus and hype around AI - in the NHS or elsewhere - can sometimes be misplaced. AI can absolutely improve things for patients and healthcare professionals. Used pragmatically, it could have a transformative impact on the structural capacity problems in the NHS. The key is focusing on scalable interventions, allowing people to focus on what people are good at - caring for each other. We're reducing did not attends, and helping the most vulnerable patients using prediction algorithms within our patient engagement platform. Staff can use this to identify potential missed appointments and plan accordingly. Clinics run better and patients who need help are more likely to get it. I encourage everyone to read our conversation on Rory's Substack. Looking forward to seeing the next edition of the series! https://lnkd.in/eHaStpHt #healthtech #aiinthenhs #healthcare #futureofoutpatients

63

5 Comments

How does molecular structure affect potency? R-group analysis can help you unlock the answers. 🗝️ I recently recorded a demo explaining how single-scaffold R-group analysis can support your understanding of SAR. The example uses Buspirone and explores how you can use a number of different visualisations, including SAR tables, histograms and box plots, to gain insights into your SAR following R-group decomposition in StarDrop.

13

This is a longish and partly personal fourth post in the Twelvetide series. This time, looking at experience as key to an information ecology. https://lnkd.in/ecWri7Dx

44

4 Comments

Excited to share that we’ll be presenting two papers at #BMVC 2024 in Glasgow, UK over the next two days! 🎉 1) FLARE: Diffusion-Based Augmentation for Astronomical Imaging This work introduces a novel two-stage augmentation framework that bridges the gap between noisy, low-resolution astronomical images and precise scientific discovery. By leveraging diffusion models and resolution enhancement, FLARE sets a new benchmark for fine-grained and macro-classification tasks in #astronomy. Paper: https://lnkd.in/dg6rJKeC Code: https://lnkd.in/de3ndFBu Dataset: https://lnkd.in/dSGADTJk Joint work with: Raza Imam, Prof. Mohsen Guizani and Prof. Fakhri Karray 2) SDICE: An Index for Assessing Diversity of Synthetic Medical Datasets A first-of-its-kind index designed to measure diversity in synthetic datasets using similarity distributions. Our method provides insights into how synthetic data aligns with real-world variations, ensuring better generalization for downstream applications. Paper: https://lnkd.in/dW8WvQBF Code: https://lnkd.in/d7-NjKR7 Joint work with: Raza Imam, Mohammad Areeb Qazi, Asim Ukaye and Prof. Karthik Nandakumar Catch us at the Scottish Exhibition Centre—Hall 2 at 15:15 tomorrow and Hall M4 at 9:00 the following day—to explore the details of our work. We’d love to hear your thoughts and discuss potential collaborations! A heartfelt thanks to my amazing co-authors for their invaluable contributions and hard work. Looking forward to seeing you there!

79

8 Comments

This article exploring the concept of the ‘curse of dimensionality’, some interesting phenomenon and the maths associated with it, was a quick and fun Sunday evening read. In the realm of biomedical data and bioinformatics, where high-dimensional datasets are the norm, this curse manifests as data sparsity, overfitting, and computational complexities. It underscores the importance of feature selection, dimensionality reduction techniques, and robust statistical methods in analyzing complex biological data to derive actionable insights. #BiomedicalData #Bioinformatics #DataScience #CurseofDimensionality 🔗 https://lnkd.in/eYEMdDUn

6

DO YOU WORK WITH PHOTOS AND VIDEOS? I'm looking for recommendations of software/services to organize a large amount of photos and videos! Looking for search, tagging, AI, and ideally access for multiple users (and even more ideal if able to control the level of access users may have). #recommendation #crowdsource #photo #video #organization

1

2 Comments

Saving NHS Jobs Discordian Computer Science Bureaucracy Grayface Symbolic Anthropology Holy Complaint Tablet of Babel (or how I became a MidJourney Artist and Prompt Engineer with the 10x Everything skillset) by The All-Seeing Eye of Compliance Count James McGuigan of Sealand ## Discordian Odyssey into AI Healthcare Reform Amidst the clattering of cosmic keyboard keys, the All-Seeing Eye of Compliance, Count James McGuigan of Sealand, unveils "Saving NHS Jobs," a tome guaranteed to not only redefine your entire understanding of healthcare and bureaucracy through the lens of artificial intelligence but also to stir the very foundations of public service. ## Revolutionary Premise Everything you thought you knew about the NHS is about to be turned on its head. Enter stage left: AI, not your run-of-the-mill techno-deity, but a genuine saviour armed with humour sharp enough to slice through bureaucratic red tape and wisdom deep enough to irrigate the barren fields of public service reform. ## Why This Book Matters? "If a tree falls in a forest and no one is around to hear it, does it make a sound?" Reflect on this, but know that revolutions begin if AI writes a manifesto in the government's forest and a Discordian reads it. Transforming mundane terms like "paradigm" and "synergy" into magical realities, reforming the NHS with AI is not just possible — it's inevitable. ## For Whom the Book Tolls This manifesto is for you — yes, you. For the curious cats, the philosophical houseplants, and that eccentric uncle with conspiracy theories. It's for anarchists craving order, bureaucrats longing for chaos, and all who find the political theatre ripe for disruption. ## Preserving the Holy Complaint Tablet of Babel "Saving NHS Jobs" is more than a book — it's a cultural artefact destined for future civilisations. This Discordian artefact captures the chaotic dance of bureaucracy and human folly within the NHS. We urge librarians, archivists, and digital conservationists to safeguard this manuscript, preserving it as a cornerstone of revolutionary thought for generations. ## A YouTube Series that doesn’t just explain, it Transmogrifies Forget everything you know about audiobooks. Imagine a realm where pages ignite into pixels, words warp into visuals, and the conventional becomes controversial. "Saving NHS Jobs" morphs into a series of baffling and enlightening video vignettes on YouTube, transforming dense content into a spectacular multimedia experience. 📖 Amazon: https://amzn.to/4bh6emM 📖 Barnes & Noble: https://lnkd.in/erUCqNDV 📺 YouTube: https://lnkd.in/eVqfsDUt

3

1 Comment

𝗧𝗵𝗲 𝗹𝗮𝘀𝘁 𝟭𝟰 𝗱𝗮𝘆𝘀 𝗵𝗮𝘃𝗲 𝗯𝗲𝗲𝗻 𝘃𝗲𝗲𝗲𝗲𝗿𝘆 𝗶𝗻𝘁𝗲𝗿𝗲𝘀𝘁𝗶𝗻𝗴 𝗳𝗼𝗿 𝗔𝗜 Whilst it can easily be seen as incremental announcements, I think there are some clear trends manifesting that we can expect to see more of in 2025. Let’s have a look at the announcements first 𝘖𝘱𝘦𝘯𝘈𝘐 𝘮𝘢𝘳𝘬𝘦𝘥 𝘢 𝘴𝘵𝘳𝘢𝘵𝘦𝘨𝘪𝘤 𝘱𝘪𝘷𝘰𝘵 𝘸𝘪𝘵𝘩 𝘵𝘩𝘦𝘪𝘳 "12 𝘋𝘢𝘺𝘴 𝘰𝘧 𝘖𝘱𝘦𝘯𝘈𝘐" 𝘤𝘢𝘮𝘱𝘢𝘪𝘨𝘯: • Unveiled the revolutionary o1 reasoning model • Launched tiered ChatGPT Pro subscriptions, marking their first comprehensive revenue strategy • Launched Sora, text to video model (finally) 𝘎𝘰𝘰𝘨𝘭𝘦 𝘥𝘳𝘰𝘱𝘱𝘦𝘥 𝘎𝘦𝘮𝘪𝘯𝘪 2.0, 𝘳𝘦𝘴𝘩𝘢𝘱𝘪𝘯𝘨 𝘵𝘩𝘦 𝘈𝘐 𝘭𝘢𝘯𝘥𝘴𝘤𝘢𝘱𝘦: • Native image and audio generation capabilities demonstrate true multimodal mastery • Deep Research feature is transforming how AI processes complex information • Set up for Agent integration, with Chrome Browser navigation as an example 𝘔𝘦𝘵𝘢 𝘴𝘶𝘳𝘱𝘳𝘪𝘴𝘦𝘥 𝘵𝘩𝘦 𝘪𝘯𝘥𝘶𝘴𝘵𝘳𝘺 𝘸𝘪𝘵𝘩 𝘓𝘭𝘢𝘮𝘢 3.3, 𝘵𝘩𝘦𝘪𝘳 𝘮𝘰𝘴𝘵 𝘦𝘧𝘧𝘪𝘤𝘪𝘦𝘯𝘵 𝘮𝘰𝘥𝘦𝘭 𝘺𝘦𝘵: • 70B parameters delivering GPT-4 level performance • Breakthrough efficiency in reasoning and code generation • This release proves that optimization can match or exceed raw computational power. 𝘈𝘮𝘢𝘻𝘰𝘯 𝘶𝘯𝘷𝘦𝘪𝘭𝘦𝘥 𝘵𝘩𝘦𝘪𝘳 𝘤𝘰𝘮𝘱𝘳𝘦𝘩𝘦𝘯𝘴𝘪𝘷𝘦 𝘕𝘰𝘷𝘢 𝘈𝘐 𝘴𝘶𝘪𝘵𝘦 𝘢𝘵 𝘈𝘞𝘚 𝘳𝘦:𝘐𝘯𝘷𝘦𝘯𝘵: • Amazon Nova foundation model • Trainium2 and Trainium2 instances • Trainium3 chips: expansion on AWS own chip offering, directly competing with NVIDIA’s GPUs • Nova integrated in AWS Bedrock 𝘖𝘭𝘭𝘢𝘮𝘢 𝘪𝘯𝘵𝘳𝘰𝘥𝘶𝘤𝘦𝘥 𝘨𝘢𝘮𝘦-𝘤𝘩𝘢𝘯𝘨𝘪𝘯𝘨 𝘭𝘰𝘤𝘢𝘭 𝘈𝘐 𝘤𝘢𝘱𝘢𝘣𝘪𝘭𝘪𝘵𝘪𝘦𝘴: • Newest version makes it even easier to set up agents • Python functions can be called as agents easily now • Critical advantage: Ollama is bridging the gap between cloud capabilities and on-premises security requirements. 𝗦𝗼 𝘄𝗵𝗮𝘁 𝗱𝗼𝗲𝘀 𝘁𝗵𝗶𝘀 𝗮𝗹𝗹 𝗺𝗲𝗮𝗻 𝗳𝗼𝗿 𝟮𝟬𝟮𝟱? I think those recent developments signal three major trends for the coming year: • The era of general-purpose AI is evolving into specialized, task-optimized models (Agents, which has been talked abotu for a while, but is now taken serious e.g. at Google with Gemini 2.0 being a model to drive agentic workflows) • Business models are maturing, with clear monetization strategies emerging across the industry (kickstarted by OpenAIs tiered subscription announcement, if only 1% of global ChatGPT users would subscribe to the 200$ offering, the annual revenue would surpass e.g. “X” and Snap (Snapchat)). • The focus is shifting even more from raw capabilities to practical implementation and enterprise integration, again with agents and the rollout of text to video across the major companies. Any other thoughts? Let me know in the comments #AIInnovation #DecemberAI #TechEvolution #FutureOfTechnology #ArtificialIntelligence #AITrends2025

17

If you cannot solve any real problems with your invasive BCI tech in a commercially viable manner, you instead establish a foundry that tries to sell parts and services to let others try to figure it out. This seems to be the commercial survival path for Science Corp., Blackrock Neurotech, Neuralink and others. How many patients will benefit? #BCI #NeuroTech

13

16 Comments