Kevin Michael Schindler
@kevinschindler.bsky.social
I'm a little torn: I can't quite decide who will come out on top:
Insane problem solvers (high IQ) who adopt AI
OR
Generalists with lots of knowledge and experience who adopt AI.
Does anyone have an opinion on this?
Insane problem solvers (high IQ) who adopt AI
OR
Generalists with lots of knowledge and experience who adopt AI.
Does anyone have an opinion on this?
June 13, 2025 at 4:05 PM
I'm a little torn: I can't quite decide who will come out on top:
Insane problem solvers (high IQ) who adopt AI
OR
Generalists with lots of knowledge and experience who adopt AI.
Does anyone have an opinion on this?
Insane problem solvers (high IQ) who adopt AI
OR
Generalists with lots of knowledge and experience who adopt AI.
Does anyone have an opinion on this?
I have never seen a more impactful UI change! 🤣🤣🤣
June 12, 2025 at 7:49 AM
I have never seen a more impactful UI change! 🤣🤣🤣
Who can relate?🤣
June 8, 2025 at 2:26 PM
Who can relate?🤣
I enjoy creating YouTube videos, but the editing process can be so time-consuming. Honestly, I have no idea how some people manage to keep up with both content production and running their company at the same time. Huge respect to them!
June 3, 2025 at 8:00 AM
I enjoy creating YouTube videos, but the editing process can be so time-consuming. Honestly, I have no idea how some people manage to keep up with both content production and running their company at the same time. Huge respect to them!
Built a $0 accountability app in 1 hour.
(It keeps me and my friends consistent.)
Here’s how it works:
• We all want to wake up before 6am
• We all want to do at least 1 pushup a day
(It keeps me and my friends consistent.)
Here’s how it works:
• We all want to wake up before 6am
• We all want to do at least 1 pushup a day
June 2, 2025 at 6:03 PM
Built a $0 accountability app in 1 hour.
(It keeps me and my friends consistent.)
Here’s how it works:
• We all want to wake up before 6am
• We all want to do at least 1 pushup a day
(It keeps me and my friends consistent.)
Here’s how it works:
• We all want to wake up before 6am
• We all want to do at least 1 pushup a day
"Nobody hires juniors anymore."
(You’ve heard it. Maybe even said it.)
But let’s be honest:
This whole “junior vs senior” debate?
It’s noise.
Here’s the truth:
→ Companies hire seniors to build out systems.
That’s it.
And here’s the twist most people miss:
(You’ve heard it. Maybe even said it.)
But let’s be honest:
This whole “junior vs senior” debate?
It’s noise.
Here’s the truth:
→ Companies hire seniors to build out systems.
That’s it.
And here’s the twist most people miss:
May 29, 2025 at 8:00 AM
"Nobody hires juniors anymore."
(You’ve heard it. Maybe even said it.)
But let’s be honest:
This whole “junior vs senior” debate?
It’s noise.
Here’s the truth:
→ Companies hire seniors to build out systems.
That’s it.
And here’s the twist most people miss:
(You’ve heard it. Maybe even said it.)
But let’s be honest:
This whole “junior vs senior” debate?
It’s noise.
Here’s the truth:
→ Companies hire seniors to build out systems.
That’s it.
And here’s the twist most people miss:
Yes, my last post was about speed vs. efficiency.
But let’s talk about who actually wins when speed becomes the strategy.
Spoiler: It’s not just the company.
It’s the consumer.
(But only when the company gets this one thing right.)
In today's AI-fueled world...
But let’s talk about who actually wins when speed becomes the strategy.
Spoiler: It’s not just the company.
It’s the consumer.
(But only when the company gets this one thing right.)
In today's AI-fueled world...
May 27, 2025 at 10:27 AM
Yes, my last post was about speed vs. efficiency.
But let’s talk about who actually wins when speed becomes the strategy.
Spoiler: It’s not just the company.
It’s the consumer.
(But only when the company gets this one thing right.)
In today's AI-fueled world...
But let’s talk about who actually wins when speed becomes the strategy.
Spoiler: It’s not just the company.
It’s the consumer.
(But only when the company gets this one thing right.)
In today's AI-fueled world...
Speed vs. Efficiency.
(This one took me years to unlearn.)
As a software and ML engineer, I was trained to worship efficiency:
No redundant loops.
No repeated code.
Optimize everything.
That mindset served me well — until it didn’t.
Now, with AI eating the world, speed is what matters.
(This one took me years to unlearn.)
As a software and ML engineer, I was trained to worship efficiency:
No redundant loops.
No repeated code.
Optimize everything.
That mindset served me well — until it didn’t.
Now, with AI eating the world, speed is what matters.
May 26, 2025 at 8:00 AM
Speed vs. Efficiency.
(This one took me years to unlearn.)
As a software and ML engineer, I was trained to worship efficiency:
No redundant loops.
No repeated code.
Optimize everything.
That mindset served me well — until it didn’t.
Now, with AI eating the world, speed is what matters.
(This one took me years to unlearn.)
As a software and ML engineer, I was trained to worship efficiency:
No redundant loops.
No repeated code.
Optimize everything.
That mindset served me well — until it didn’t.
Now, with AI eating the world, speed is what matters.
Guys, I need a break please. Here is the tldr.
🧠 New Models: Opus 4 & Sonnet 4
1️⃣Claude Opus 4:
- Best coding model worldwide (SWE-bench 72.5%, Terminal-bench 43.2%).
- Sustains long tasks and complex workflows over hours.
- Significantly improved memory when given file access.
🧠 New Models: Opus 4 & Sonnet 4
1️⃣Claude Opus 4:
- Best coding model worldwide (SWE-bench 72.5%, Terminal-bench 43.2%).
- Sustains long tasks and complex workflows over hours.
- Significantly improved memory when given file access.
May 22, 2025 at 8:14 PM
Guys, I need a break please. Here is the tldr.
🧠 New Models: Opus 4 & Sonnet 4
1️⃣Claude Opus 4:
- Best coding model worldwide (SWE-bench 72.5%, Terminal-bench 43.2%).
- Sustains long tasks and complex workflows over hours.
- Significantly improved memory when given file access.
🧠 New Models: Opus 4 & Sonnet 4
1️⃣Claude Opus 4:
- Best coding model worldwide (SWE-bench 72.5%, Terminal-bench 43.2%).
- Sustains long tasks and complex workflows over hours.
- Significantly improved memory when given file access.
Wrote a script to find invoices in my Gmail account and save them.
Next stop: Using MCP to log in to my accounts and download them automatically when needed.
All done in a couple of minutes with some prompts and the right tools.
What a time to be alive.
Next stop: Using MCP to log in to my accounts and download them automatically when needed.
All done in a couple of minutes with some prompts and the right tools.
What a time to be alive.
May 10, 2025 at 3:46 PM
Wrote a script to find invoices in my Gmail account and save them.
Next stop: Using MCP to log in to my accounts and download them automatically when needed.
All done in a couple of minutes with some prompts and the right tools.
What a time to be alive.
Next stop: Using MCP to log in to my accounts and download them automatically when needed.
All done in a couple of minutes with some prompts and the right tools.
What a time to be alive.
🔬 Machine Learning in Biotech Upstream Development:
(Highlighting what is currently being used and the results you can expect.)
✅ML-Guided Cyclodextrin Selection for Terpene Biosynthesis
Study: S. Cai et al., J. Agric. Food Chem., 2025
Objective: Improve production and capture of volatile terpenes
(Highlighting what is currently being used and the results you can expect.)
✅ML-Guided Cyclodextrin Selection for Terpene Biosynthesis
Study: S. Cai et al., J. Agric. Food Chem., 2025
Objective: Improve production and capture of volatile terpenes
March 24, 2025 at 9:00 AM
🔬 Machine Learning in Biotech Upstream Development:
(Highlighting what is currently being used and the results you can expect.)
✅ML-Guided Cyclodextrin Selection for Terpene Biosynthesis
Study: S. Cai et al., J. Agric. Food Chem., 2025
Objective: Improve production and capture of volatile terpenes
(Highlighting what is currently being used and the results you can expect.)
✅ML-Guided Cyclodextrin Selection for Terpene Biosynthesis
Study: S. Cai et al., J. Agric. Food Chem., 2025
Objective: Improve production and capture of volatile terpenes
Favorite DoE (Design of Experiments) method?
March 8, 2025 at 7:22 PM
Favorite DoE (Design of Experiments) method?
From startup to enterprise—these fundamentals stay the same.
If you're working with AI, whether in a small project or a large organization, some core technologies remain constant.
Here’s what you should focus on:
If you're working with AI, whether in a small project or a large organization, some core technologies remain constant.
Here’s what you should focus on:
March 8, 2025 at 10:00 AM
From startup to enterprise—these fundamentals stay the same.
If you're working with AI, whether in a small project or a large organization, some core technologies remain constant.
Here’s what you should focus on:
If you're working with AI, whether in a small project or a large organization, some core technologies remain constant.
Here’s what you should focus on:
Ranking Machine Learning Models for Titer Increase in Biotech:
1. XGBoost: Often performs best due to its advanced boosting techniques, handling nonlinear relationships well, and offering high accuracy in predictions.
1. XGBoost: Often performs best due to its advanced boosting techniques, handling nonlinear relationships well, and offering high accuracy in predictions.
March 4, 2025 at 8:57 AM
Ranking Machine Learning Models for Titer Increase in Biotech:
1. XGBoost: Often performs best due to its advanced boosting techniques, handling nonlinear relationships well, and offering high accuracy in predictions.
1. XGBoost: Often performs best due to its advanced boosting techniques, handling nonlinear relationships well, and offering high accuracy in predictions.
🔬 Predicting drug quality before it’s too late?
Biopharma manufacturers usually wait days for lab tests to confirm if a batch is good. But what if we could predict potency 5 days in advance?
A new study by Bayer Pharmaceuticals uses machine learning to do just that. 🚀
Biopharma manufacturers usually wait days for lab tests to confirm if a batch is good. But what if we could predict potency 5 days in advance?
A new study by Bayer Pharmaceuticals uses machine learning to do just that. 🚀
March 3, 2025 at 6:29 PM
🔬 Predicting drug quality before it’s too late?
Biopharma manufacturers usually wait days for lab tests to confirm if a batch is good. But what if we could predict potency 5 days in advance?
A new study by Bayer Pharmaceuticals uses machine learning to do just that. 🚀
Biopharma manufacturers usually wait days for lab tests to confirm if a batch is good. But what if we could predict potency 5 days in advance?
A new study by Bayer Pharmaceuticals uses machine learning to do just that. 🚀
👩🔬 AI Will Not Replace Scientists – It Will Work With Them
(Here’s what that looks like.)
Google just launched the AI Co-Scientist.
But it’s not about replacing researchers.
(Here’s what that looks like.)
Google just launched the AI Co-Scientist.
But it’s not about replacing researchers.
Google AI co-scientist Trusted Tester Program
Thank you for your Organization's interest in the Google AI co-scientist Trusted Tester Program. Please fill out the following form to express your Organization's interest and provide us with more…
buff.ly
February 21, 2025 at 4:35 PM
👩🔬 AI Will Not Replace Scientists – It Will Work With Them
(Here’s what that looks like.)
Google just launched the AI Co-Scientist.
But it’s not about replacing researchers.
(Here’s what that looks like.)
Google just launched the AI Co-Scientist.
But it’s not about replacing researchers.
🚀 AI Co-Scientist Outperforms SOTA Models on 15 Research Problems
(And you could get early access.)
Google’s AI Co-Scientist just beat leading agentic and reasoning models across 15 complex research tasks.
(And you could get early access.)
Google’s AI Co-Scientist just beat leading agentic and reasoning models across 15 complex research tasks.
Google AI co-scientist Trusted Tester Program
Thank you for your Organization's interest in the Google AI co-scientist Trusted Tester Program. Please fill out the following form to express your Organization's interest and provide us with more…
buff.ly
February 21, 2025 at 8:57 AM
🚀 AI Co-Scientist Outperforms SOTA Models on 15 Research Problems
(And you could get early access.)
Google’s AI Co-Scientist just beat leading agentic and reasoning models across 15 complex research tasks.
(And you could get early access.)
Google’s AI Co-Scientist just beat leading agentic and reasoning models across 15 complex research tasks.
Google Introduces the AI Co-Scientist – A Multi-Agent System Built on Gemini 2.0
A virtual research collaborator designed to:
🔬 Generate novel scientific hypotheses
🔬 Debate and refine ideas using self-play and tournaments
🔬 Accelerate biomedical discoveries alongside human experts
A virtual research collaborator designed to:
🔬 Generate novel scientific hypotheses
🔬 Debate and refine ideas using self-play and tournaments
🔬 Accelerate biomedical discoveries alongside human experts
Google AI co-scientist Trusted Tester Program
Thank you for your Organization's interest in the Google AI co-scientist Trusted Tester Program. Please fill out the following form to express your Organization's interest and provide us with more…
docs.google.com
February 20, 2025 at 4:35 PM
Google Introduces the AI Co-Scientist – A Multi-Agent System Built on Gemini 2.0
A virtual research collaborator designed to:
🔬 Generate novel scientific hypotheses
🔬 Debate and refine ideas using self-play and tournaments
🔬 Accelerate biomedical discoveries alongside human experts
A virtual research collaborator designed to:
🔬 Generate novel scientific hypotheses
🔬 Debate and refine ideas using self-play and tournaments
🔬 Accelerate biomedical discoveries alongside human experts
Deep Learning is Transforming Medical Imaging.
(Clinicians, data scientists, AI researchers—this one's for you.)
Over the past decade, deep learning has revolutionized medical image analysis.
(Clinicians, data scientists, AI researchers—this one's for you.)
Over the past decade, deep learning has revolutionized medical image analysis.
February 18, 2025 at 4:35 PM
Deep Learning is Transforming Medical Imaging.
(Clinicians, data scientists, AI researchers—this one's for you.)
Over the past decade, deep learning has revolutionized medical image analysis.
(Clinicians, data scientists, AI researchers—this one's for you.)
Over the past decade, deep learning has revolutionized medical image analysis.
🚀 Machine Learning is transforming biopharmaceutical development and manufacturing like never before.
Here are some key trends shaping the future:
Here are some key trends shaping the future:
February 10, 2025 at 8:57 AM
🚀 Machine Learning is transforming biopharmaceutical development and manufacturing like never before.
Here are some key trends shaping the future:
Here are some key trends shaping the future:
Here’s how Multi-Modal Learning can drastically improve the prediction of differential transcript expression by combining DNA, RNA, and protein data:
InstaDeepAI/isoformer · Hugging Face
We’re on a journey to advance and democratize artificial intelligence through open source and open science.
huggingface.co
February 7, 2025 at 8:57 AM
Here’s how Multi-Modal Learning can drastically improve the prediction of differential transcript expression by combining DNA, RNA, and protein data:
Reposted by Kevin Michael Schindler
🚀 Find the Best Ollama LLM in 1 Minute!
1️⃣ Visit 👉 Hugging Face Leaderboard huggingface.co/spaces/open-...
2️⃣ Open "Advanced Filters"
3️⃣ Check ✅ "Only Official Providers"
4️⃣ Set the model size slider:
Rule of thumb: 📏 VRAM x 1.5 = Model Size (in B)
1️⃣ Visit 👉 Hugging Face Leaderboard huggingface.co/spaces/open-...
2️⃣ Open "Advanced Filters"
3️⃣ Check ✅ "Only Official Providers"
4️⃣ Set the model size slider:
Rule of thumb: 📏 VRAM x 1.5 = Model Size (in B)
Ollama
Get up and running with large language models.
ollama.com
January 4, 2025 at 11:24 AM
🚀 Find the Best Ollama LLM in 1 Minute!
1️⃣ Visit 👉 Hugging Face Leaderboard huggingface.co/spaces/open-...
2️⃣ Open "Advanced Filters"
3️⃣ Check ✅ "Only Official Providers"
4️⃣ Set the model size slider:
Rule of thumb: 📏 VRAM x 1.5 = Model Size (in B)
1️⃣ Visit 👉 Hugging Face Leaderboard huggingface.co/spaces/open-...
2️⃣ Open "Advanced Filters"
3️⃣ Check ✅ "Only Official Providers"
4️⃣ Set the model size slider:
Rule of thumb: 📏 VRAM x 1.5 = Model Size (in B)
I'm thinking about joining the precisionFDA GenAI Community Challenge.
I’ve already built and tested many RAG pipelines, so this feels like the perfect opportunity to put them to the test. ⬇️
I’ve already built and tested many RAG pipelines, so this feels like the perfect opportunity to put them to the test. ⬇️
precisionFDA
A community for NGS assay evaluation
precision.fda.gov
January 4, 2025 at 11:31 AM
I'm thinking about joining the precisionFDA GenAI Community Challenge.
I’ve already built and tested many RAG pipelines, so this feels like the perfect opportunity to put them to the test. ⬇️
I’ve already built and tested many RAG pipelines, so this feels like the perfect opportunity to put them to the test. ⬇️
Reposted by Kevin Michael Schindler
Opportunity in #peptide chemistry at AstraZeneca (Cambridge UK) ref: R-215639 | closing: 13-Jan-2025 #SynChem #UKChemJobs #ChemJobs #chemsky 🧪
astrazeneca.wd3.myworkdayjobs.com/Careers/job/...
astrazeneca.wd3.myworkdayjobs.com/Careers/job/...
Senior Scientist – Peptide Chemistry
Job Title: Senior Scientist – Peptide Chemistry Location: Cambridge Salary: competitive Introduction to role: AstraZeneca is a global, science-led, patient-focused biopharmaceutical company that focus...
astrazeneca.wd3.myworkdayjobs.com
January 3, 2025 at 10:16 AM
Opportunity in #peptide chemistry at AstraZeneca (Cambridge UK) ref: R-215639 | closing: 13-Jan-2025 #SynChem #UKChemJobs #ChemJobs #chemsky 🧪
astrazeneca.wd3.myworkdayjobs.com/Careers/job/...
astrazeneca.wd3.myworkdayjobs.com/Careers/job/...
🚀 Find the Best Ollama LLM in 1 Minute!
1️⃣ Visit 👉 Hugging Face Leaderboard huggingface.co/spaces/open-...
2️⃣ Open "Advanced Filters"
3️⃣ Check ✅ "Only Official Providers"
4️⃣ Set the model size slider:
Rule of thumb: 📏 VRAM x 1.5 = Model Size (in B)
1️⃣ Visit 👉 Hugging Face Leaderboard huggingface.co/spaces/open-...
2️⃣ Open "Advanced Filters"
3️⃣ Check ✅ "Only Official Providers"
4️⃣ Set the model size slider:
Rule of thumb: 📏 VRAM x 1.5 = Model Size (in B)
Ollama
Get up and running with large language models.
ollama.com
January 4, 2025 at 11:24 AM
🚀 Find the Best Ollama LLM in 1 Minute!
1️⃣ Visit 👉 Hugging Face Leaderboard huggingface.co/spaces/open-...
2️⃣ Open "Advanced Filters"
3️⃣ Check ✅ "Only Official Providers"
4️⃣ Set the model size slider:
Rule of thumb: 📏 VRAM x 1.5 = Model Size (in B)
1️⃣ Visit 👉 Hugging Face Leaderboard huggingface.co/spaces/open-...
2️⃣ Open "Advanced Filters"
3️⃣ Check ✅ "Only Official Providers"
4️⃣ Set the model size slider:
Rule of thumb: 📏 VRAM x 1.5 = Model Size (in B)