Jacob Glanville

AI survival tip #1: instead of prompting AI systems to perform analysis, prompt them to write code to perform the analysis, then to run the code and provide both the code and the results as outputs. Solves the traceability and logic transferability problem.

I have been reflecting on this for years, and have decided to make the position public. I think there should be an international band on the serial passage of zoonotic (non-human) viruses in human cell lines. Groups may do this for reasons of convenience to adapt the virus to culture more easily in the lab, but the practice risks adapting the virus to human infectivity (anti-attenuation). A non-human cell line (MDCK etc) should be used instead for this purpose. It’s not a perfect defense, but adapting viruses to propagate more easily in human cell lines is an insanely bad example of penny wise and pound foolish lab practice.

This Russian Anti-Plague Institute death may have been a bioweapon exposure. Many antibiotics work on plagues. Surely she would have been given them after the vial broke. So would the 200 contacts. This suggests a multi-drug resistant form not found in nature.

Furthermore, the state response doesn’t line up well with the facts. She broke a vial containing a rare plague at the anti-Plague institute, then a few days later got sick and died very rapidly. She’s in her late 20s: this is not a normal pneumonia progression except very normal for plague. Early reports were that it was plague - only later government reports deny it was a pathogen work related, but that answer fits so poorly with the facts, and so well with a coverup of a bioweapon leak, that I think we should look deeper. This isn’t proof, but it’s a lot of circumstantial evidence. The world must say no to bioweapons programs.

People say that AI means you can’t trust any images anymore. Maybe this is liberating: we never should have trusted distributed images that were curated and orchestrated to advertise, propagandize, and frame our worldview. AI didnt invent deceptive imagery - it just industrialized it. Maybe there’s a freedom in the internet being so hopelessly polluted that people will return to the dignity and common sense of what they can see around them with their own eyes.

12% of global cancer attributable to infectious disease in this study. The number is likely a significant undercount because it only includes a few pathogens where the relationship has been established. It should be a central objective of the coming century to eradicate as many pathogens as possible. t.co/u412yZUhpV

@therealRYC The mind blowing thing is how relatively functional the negative controls were who had their cortex missing and yet weren’t catatonic

@nutrition_facts Mortality remains at 100% in both cases

@anthonystaj @nytimes 100% I was just blocked yesterday for asking for a research survey into some very standard vaccine questions

@yashar @andersoncooper @hilbertspaess Harry Potter is a real downer

@RuxandraTeslo @IFP @kroetscha Our Phase Ia was in Australia. Great partners.

@chess_feed Rc7,BxR,KnxB

@iskander Let’s talk

@justpaulemerson

@christopher3700 @SamaHoole We are in clinical trials

I believe in the medical utopia but it’s probably will take 2 centuries of further steady work to get there. A reminder to folks on this thread that it takes 10-20 years of a medical concept to go from concept to approval. AI as a modality currently has a 0% success rate and some notable failures. Give it time to contribute to medicine. “In the next 20 years we will see meaningful medical advances brought about by AI” is a credible statement.

Seeing a lot of instant gratification responses to Florian’s comment. I just want to politely point out that it takes about 10-20 years for a concept to go from early bench to approval. AI-generated X as a modality currently has a 0% success track record and some notable failures. I’m not betting against innovation or other people’s work, but everyone wants an instant gratification fix to disease. It takes time to do it right. My estimate is that we will be at this for at least 2 centuries. But I also do think that a medical eutopia is eventually possible. It just will take time.

@SamaHoole However, you’ve got a good list. Medical science has done more good than bad but clearly there have been screw ups along the way. Taking birthing away from midwives and making woman lie on their backs and escalate through a whole stack of medicines and interventions is another

I recommend you drop the tone and spend a little more time educating yourself. Check out the history of the programs in Australia, Germany, Hawaii, and Denmark as some examples of successful high penetration programs. Then reread my original comment and we can have a useful conversation.

While this is excellent progress, California continues to have some of the highest electricity bills in the country. We need national legislation that protects the ability of individual homeowners to install solar and reap equitable benefits from selling their surplus back to the grid. Such legislation would spontaneously move the whole country towards renewable energy through self interest while also helping a lot of people become “micro business” energy producers. The rules right now are terrible for single home contributers to sell power into the grid.

@SciGuySpace Second question - for rapid reuse, what stops them from just applying a new heat shield every time? Surely the heat shield isn’t the majority of the cost of the rocket

Glanville, whose company is currently working to develop a universal flu vaccine, among others, summed up the impact of these and other vaccines in America. "To truly understand the impact that vaccines have had in the last 100 years, read through the list of pathogens along the timeline until they become familiar to you, until they include things you or your family may have been infected by in your own life," he advised. "You may barely even recognize most of the pathogens in the first half of the list – because vaccines pushed them out of the human experience. That is the power of vaccines." t.co/0QbSJyTV9K

@DrNeilStone You are right - this is going to be a shitshow. It’s containable but we need to act agressively. The situation on the ground is just a mess and the international support infrastructure that we had back in the 2014 outbreak just isn’t there now.

Looks like other research groups are now updating to align with my estimates that I posted on the 17th. abcnews.com/amp/Internatio…

I had previously reported my estimate of 2.5 months when most of media was reporting a month of outbreak (and therefore coming to an incorrect conclusion about transmission rate). It looks like others are now coming to the same conclusion as my estimates : abcnews.com/amp/Internatio…