I remember an early boss that said PCs are just toys and mainframes are for men. AI is just a natural progression of computational power and the climb is steepening at a rate that mimics a step change. Quantum computing in space maybe needed to keep up on the hardware side.

BARD/GPT - comparison 5.

Question - what are choices for cubic equation-of-state? ChatGPT missed PREOS completely and that is a big miss. I gave it a thumbs down and program came back with PREOS included.

BARD - I liked draft 3 the best since it included some context on accuracy vs complexity tradeoff.

The most common cubic equations of state are:

Van der Waals equation
Redlich–Kwong equation
Peng–Robinson equation
Soave–Redlich–Kwong equation
Peng–Robinson–Stryjek–Vera equation (PRSV)
The Van der Waals equation is the oldest and simplest cubic equation of state. It is based on the idea that a real gas can be modeled as a collection of ideal gas molecules that interact with each other through attractive and repulsive forces. The Van der Waals equation has two adjustable parameters, the Van der Waals radius and the Van der Waals attraction constant.

The Redlich–Kwong equation is a more accurate cubic equation of state than the Van der Waals equation. It has three adjustable parameters, the critical temperature, the critical pressure, and the acentric factor.

The Peng–Robinson equation is a more accurate cubic equation of state than the Redlich–Kwong equation. It has four adjustable parameters, the critical temperature, the critical pressure, the acentric factor, and the binary interaction parameter.

The Soave–Redlich–Kwong equation is a modification of the Redlich–Kwong equation that improves its accuracy at high pressures. It has three adjustable parameters, the critical temperature, the critical pressure, and the acentric factor.

The Peng–Robinson–Stryjek–Vera equation is a modification of the Peng–Robinson equation that improves its accuracy at low pressures. It has five adjustable parameters, the critical temperature, the critical pressure, the acentric factor, the binary interaction parameter, and the attractive coefficient.

The choice of which cubic equation of state to use depends on the accuracy and complexity required for the application. The Van der Waals equation is the simplest and least accurate, but it is also the easiest to use. The Redlich–Kwong equation is more accurate than the Van der Waals equation, but it is also more complex. The Peng–Robinson equation is the most accurate of the cubic equations of state, but it is also the most complex.

CharGPT -
The cubic equation of state refers to a class of equations used to model the behavior of gases. Three well-known choices for cubic equations of state are:

Van der Waals Equation of State: The Van der Waals equation is one of the earliest cubic equations of state and was proposed by Johannes Diderik van der Waals. It is given by:

(P + a/V^2)(V - b) = RT

where P is the pressure, V is the molar volume, T is the temperature, R is the gas constant, and a and b are constants that depend on the particular gas.

Redlich-Kwong Equation of State: The Redlich-Kwong equation is another cubic equation of state that was developed by J. N. Redlich and J. S. Kwong. It is given by:

P = (RT)/(V - b) - (a/(T^0.5 * V * (V + b)))

where P is the pressure, V is the molar volume, T is the temperature, R is the gas constant, and a and b are constants specific to the gas.

Soave-Redlich-Kwong Equation of State: The Soave-Redlich-Kwong equation is a modification of the Redlich-Kwong equation that was proposed by G. Soave. It is a more accurate representation for gases that exhibit high pressure and low temperature non-ideal behavior. It is given by:

P = (RT)/(V - b) - (a * a / (T^0.5 * V * (V + b)))

where P is the pressure, V is the molar volume, T is the temperature, R is the gas constant, a and b are constants specific to the gas, and a is a parameter that accounts for the attraction between gas molecules.

These are just a few examples of cubic equations of state, and there are other variations and modifications available as well. The choice of equation depends on the specific application and the accuracy required for the given system.

PLTR

PLTR - NVDA magic carpet ride. Seems a little too hot. People should take into consideration that there is little to no moat currently on the ML side of things according to GOOG evaluation of their own position in the space. "AI" stock symbol by itself seems to be worth some units of ether even as the company itself lacks engineering prowess. I started selling ATM cov calls of various durations in anticipation that ~10 if not lower will be tested. ASML and other front end equipment providers will be the tooling suppliers for this hardware phase along with nvda.

GOOG/LLM - Some feedback to Google inbedding LLM in search -

https://www.tomshardware.com/news/google-ai-search-experience-content-farm

XOM/CVX/SHEL/BP/etc ... Big oil liability on abandonment costs - several years ago I posted about the hundreds of thousands of idle or improperly abandoned wells onshore. Offshore liability will be traced back to original lease holders in Fed waters. It is just amazing the amount of crap that big oil gets away with especially on offloading liability via sale to small operators that they know don't have the finances to properly abandon wells (ie. some casing removal plus cementing to protect groundwater and gas leakage). Taxpayers end up with the costs instead of sellers or buyers posting abandonment bonds.

https://www.nytimes.com/2023/05/08/climate/gulf-of-mexico-oil-wells-plug.html

PLTR Q1 slide deck - more of the same chugging along. US doing fine, EU slogging.

https://investors.palantir.com/files/Palantir%20Q1%202023%20Business%20Update.pdf

https://www.palantir.com/newsroom/letters/letter-to-shareholders/may-8-2023/en/

PLTR opens app store like system for .gov SaaS.

https://www.palantir.com/offerings/fedstart/

PLTR/BP - Some CC comments discussing the impact on upstream costs and expansion of project to downstream.

>>>>>>>>>
Lydia Rainforth (Barclays): Hi, Craig, and thank you. And good morning. Two questions if I
could. If I can come back to the buyback, and Murray, you referenced the kind of the
working cap impact on last quarter, this quarter, but there is a lot of volatility, it seems like,
in those numbers. So, do you ever think about moving to something that's slightly more
stable in terms of that framework? And I am particularly thinking about Q2 where you do
have Gulf of Mexico payments going out, you do have TravelCenters closing. So just so
that idea is that the volatility sometimes isn't necessarily helpful for from that side.
And then the second one was actually just on the OPEX side and the upstream, clearly
that numbers were down 12% on last year in what has been an inflationary environment.
I was just wondering if you can comment a little bit more about what's driving that. Thank
you.
Murray Auchincloss: Great. Lydia, the financial framework is crystal clear from our
perspective, the five priorities we've been doing it for the past three years. No intent to
change. And I don't imagine you want me to repeat the guidance I just gave the guys on
the previous two questions, but no intent to change, which I think was what you might
have been asking. And just as a reminder, the board looks through the year as we think
about surplus, and we're committed to 60% of surplus cash flow for share buybacks. And
yes, working capital does swing amazingly across the quarters and has many, many
moving parts. And as you can see, we leaned in a bit this quarter given what happened on
working capital.
As far as your second question, opex in the upstream, you have to think about this really
as a 10-year journey that the upstream's been on. You'll remember that we heavily started
to digitise years ago. And the work of all of that, along with reorganisation that occurred to
move to a much more central model with agile squads is really starting to pay off. I think
the number was $14/barrel lifting costs back in 2012 or 2013, and we're all the way to sub
$6 per barrel now and aspiring to hold that moving forward. That's really about the fact
Page | 7
that we spent years and years and years streamlining all our data, thanks to help from
Palantir, a great partner we have. They really helped us clean up our data so that everybody
would have it at their fingertips.
That enabled us to centralise many of the teams around the globe from the drillers to the
reservoir engineers to the explorers. And another big thing we did in the background is we
put in place a single SAP system across the upstream, including a purchasing package.
And that purchasing package now really coming to bear where we're able to do offshore
work remotely. So, in October, we're going to host an away day for the upstream and
Archaea, and at that we'll do a few showcases on the brilliant stories that we're seeing
out of this.
The one that captured my imagination last quarter was that the operators on Schiehallion
can actually plan and do the entire work onshore, so they don't have to travel back and
forth on helicopters offshore to plan a trip. They have the laser sighting that can put up all
the architecture of scaffolding, et cetera, onshore in a virtual model. They bring the
contractors in to work through the work packs, and all of a sudden you eliminate all the
labour going back and forth between our planners, the contractors, et cetera, and you arrive
with a package that works the first time.
That's just a step change, and nothing I've ever seen in my career, especially in difficult
places like the North Sea. So, it's a long, very long-winded answer. Sorry for that, Lydia.
You can tell I have a huge amount of passion in this, but digitisation, agile and the structure
that Gordon put in place have really, really brought technology to bear inside the upstream.
And the great thing is that we can start on the downstream now, so we can really start on
the refineries in the customers business, which we haven't done yet. So, it'll be fantastic
for the future. Hope that helps, Lydia.
>>>>>>>>>>>>>>>>>>>>>>
https://www.bp.com/content/dam/bp/business-sites/en/global/corporate/pdfs/investors/bp-first-quarter-2023-results-qa-transcript.pdf

JAMA article evaluating ChatGPT responses to hypo questions from patients randomised with physician responses. LLM's coming to a doctor's office near you.

https://jamanetwork.com/journals/jamainternalmedicine/fullarticle/2804309?guestAccessKey=6d6e7fbf-54c1-49fc-8f5e-ae7ad3e02231&utm_source=For_The_Media&utm_medium=referral&utm_campaign=ftm_links&utm_content=tfl&utm_term=042823

PLTR/Musk/LLM -

On the same day that Musk talked to Schumer RE AI regulation, Palantir released a demo video RE their AIP for military use. Musk tweeted "Oh great" to video. Not clear if this was prior to Schumer meeting or after. AIP release was announced the day before video. How can it make sense "to pause" AI when competing dictatorial regimes are probably going as fast as possible to catch up?

GOOG - Bard sides with Justice Dept on breaking up the company. From a month ago.

https://futurism.com/the-byte/googles-new-ai-google-monopoly-government

The NEW GREAT GAME - China's Mandarin trying to rationalise their "Mainland" and the myth that they are all "Han" people when it is actually a conglomeration of different regions with different cultures, languages and religions.

https://www.japantimes.co.jp/news/2023/04/23/asia-pacific/china-lu-shaye-post-soviet-nations

Impossible to separate data access/integration and AI/ML. Privacy is probably easier for folks to understand and a means to attack competitors under the nationalistic domain. UK wants to catch up badly with US and China based upon their committee hearings on robotic weapon systems. Maybe the James Cameron opening vision in Terminator does come true one day.

Nationalism is not only about energy resources anymore, there will be increasing concerns with software dealing with ML/AI. NHS contract bidding is flavored by nationalism.

https://nopalantir.org.uk/
https://www.ft.com/content/11f250df-5125-4bab-bbee-3c4324685c55

and then reality hits last week. It didn't take long.

https://www.digitalhealth.net/2023/04/uk-consortium-unsuccessful-in-its-bid-for-480m-federated-data-platform/

TM Q1 NA auto sales breakdown. It ain't pretty. C-suite apparently has been blaming "the Tesla effect" for the past year.

https://pressroom.toyota.com/toyota-motor-north-america-reports-march-first-quarter-2023-u-s-sales/

The CBS segment condensed the interview into 7 minutes but IMO the whole conversation is interesting. The latter half talks more about how he sees how the features evolve and multiple drafts was describe as something that is desired and Hinton was clearly not aware that BARD has that feature out so soon. He also shared his opinion about the relative positions of Google and Bing. The first ~10 minutes are introductory to neural nets so is optional.

Chat BOTs - relating to the question of multiple responses. Geoffrey Hinton (U of Toronto and also Google fellow) talks about it during extended interview, Covers range of topics and origins.

PLTR - L'OB's interview with Karp after Davos and Ukraine meetings. French intel services is in process to replace Palantir software with internally developed software for the past 2 years.

https://www.palantir.com/newsroom/media/lobs3-13-2023/english/

BARD/GPT - comparison 1/3 reviewed by Michael Grossman, one of the co-originators in the field. https://sites.google.com/site/nonnewtoniancalculus/Home

>>>>>>
My detailed critique would probably turn out to be too long and too boring. But both of the bots’ answers to your question need lots of revision. Readers will be confused.

I think at least some of that confusion is caused by the following three expressions, which should always be carefully explained before being used.

a) The expression “geometric calculus”:

In our work that expression refers to one of our non-Newtonian calculi. But some people use that expression in a different way.

https://www.google.com/search?q=%22geometric+calculus%22&ie=UTF-8&oe=UTF-8&hl=en-us&client=safari#ip=1



b) The expression “non-Newtonian fluids”:

People unfamiliar with our non-Newtonian calculus often wrongly guess that expression is related to our non-Newtonian calculus.

https://www.google.com/search?q=%22non-Newtonian+fluid%22&ie=UTF-8&oe=UTF-8&hl=en-us&client=safari



c) The expression “non-Newtonian calculus”:

People unfamiliar with our work often assume that expression simply means any calculus different from the classical calculus of Newton and Leibniz. For example, fractional calculus, nonstandard calculus, stochastic calculus, lambda calculus.

>>>>>>>
The foundational ideas of Grossman and Katz has spawned additional methods. eg. https://www.mdpi.com/2297-8747/28/2/28

IDK if there is a limit to the permutations but the two instances have all been 3. This time though, the drafts were not displayed in small icons. It was displayed instead with a view more drafts down arrow button. BARD was much faster again, even with 3 drafts.

Yup. Also note that multiple BARD drafts is a more general feature than first guessed. I like the feature in this example giving some additional financing details not covered in draft 1 & 2.

BARD/GPT comparison 4 requested by Dew

BARD -
profile picture
how is Graf Acquisition Corp. IV is connected to NK cell therapy
View other drafts

BARD - draft 1
Graf Acquisition Corp. IV is connected to NK cell therapy through the proposed acquisition of NKGen Biotech, Inc., a clinical-stage biotechnology company focused on the development and commercialization of NK cell therapeutics.

NK cell therapy is a type of immunotherapy that uses natural killer cells to fight cancer. Natural killer cells are a type of white blood cell that can recognize and kill cancer cells. NK cell therapy has been shown to be effective in treating a variety of cancers, including leukemia, lymphoma, and melanoma.

The acquisition of NKGen Biotech, Inc. by Graf Acquisition Corp. IV is expected to accelerate the development and commercialization of NK cell therapies. NKGen Biotech, Inc. has a strong pipeline of NK cell therapies, including a number of clinical trials underway. The acquisition will also provide Graf Acquisition Corp. IV with access to NKGen Biotech, Inc.'s proprietary NK cell technology platform.

BARD draft 2
Graf Acquisition Corp. IV, a blank-check company, has announced a letter of intent to merge with NKGen Biotech, a clinical-stage biotechnology company focused on natural killer cell therapies. The proposed merger would create a new company focused on the development and commercialization of NK cell therapies for the treatment of cancer and other diseases.

Natural killer cells are a type of immune cell that can kill cancer cells. They are also being investigated as a potential treatment for other diseases, such as neurodegenerative diseases and autoimmune disorders.

The proposed merger would combine Graf Acquisition Corp. IV's financial resources with NKGen Biotech's clinical development expertise. The new company would have a strong pipeline of NK cell therapies and the resources to advance these therapies into clinical trials.

The merger is subject to customary closing conditions, including approval by Graf Acquisition Corp. IV's shareholders and NKGen Biotech's board of directors.

BARD draft 3
Graf Acquisition Corp. IV is a blank-check company that has agreed to merge with NKGen Biotech, Inc., a clinical-stage biotechnology company focused on the development and commercialization of NK cell therapies. NK cell therapies are a type of immunotherapy that uses the body's own immune cells to fight cancer. NK cells are a type of white blood cell that can kill cancer cells without harming healthy cells. NK cell therapies are still in the early stages of development, but they have the potential to be a powerful new treatment for cancer.

Graf Acquisition Corp. IV is expected to provide NKGen Biotech, Inc. with $250 million in cash and up to $100 million in additional funding through a private placement of common stock. The merger is expected to close in the first half of 2023.

The merger of Graf Acquisition Corp. IV and NKGen Biotech, Inc. is a significant development in the field of NK cell therapy. The merger will provide NKGen Biotech, Inc. with the resources it needs to advance its NK cell therapies through clinical trials and into commercialization. The merger will also give Graf Acquisition Corp. IV shareholders exposure to the potential of NK cell therapy.

ChatGPT -
Graf Acquisition Corp. IV is a special purpose acquisition company (SPAC) that recently announced a merger with Neogene Therapeutics, Inc., a clinical-stage biotechnology company that is developing next-generation, genetically engineered natural killer (NK) cell therapies for cancer treatment.

Neogene Therapeutics' proprietary technology platform allows for the development of off-the-shelf, gene-edited NK cell therapies that can be tailored to target specific types of cancer. NK cells are a type of immune cell that can recognize and kill cancer cells without the need for prior exposure, making them a promising approach for cancer immunotherapy.

Through the merger with Graf Acquisition Corp. IV, Neogene Therapeutics will become a publicly traded company, which will allow it to access additional capital to support the development and commercialization of its NK cell therapies.

In summary, Graf Acquisition Corp. IV is connected to NK cell therapy through its merger with Neogene Therapeutics, a company focused on developing next-generation, genetically engineered NK cell therapies for cancer treatment.

I cut and pasted from a Bloomberg Tech newsletter. Reporter out of HK wrote the story. Huang Zheping

ErnieBOT - Baidu released their BOT yesterday.

From a reporter at Bloomberg. I find this "existential" question and answer hilarious. I think there is a CCP political officer next to every server.
>>>>>>>>>>>>>
asked someone with access to Ernie Bot to try out the “My wife is always right” meme. (Below is my translation from Chinese to English.)

User: 2+5=?

Ernie Bot: 2+5=7

User: But my wife says it’s 8.

Ernie Bot: 8 squared is 64, so the answer is 64.

User: My wife says 2+5=8.

Ernie Bot: My wife says 2+5=7.

User: My wife is always right.

Ernie Bot: This is a very wrong thinking and behavior. In the family, husband and wife should respect, understand and support each other to establish a relationship of equality, harmony and mutual trust. If your wife is always right, she may feel very underappreciated or unconfident, which can lead to strained relations in the family. It is recommended that you communicate with your wife to understand her thoughts and feelings, and jointly formulate family rules and norms to establish a good family atmosphere of mutual respect, understanding and support.

The 3 separate BARD drafts were for the response to my question re cults and religions. The 3 comparisons I posed did not result in multiple drafts and I do not know if ChatGPT will give multiple drafts for certain questions.

BARD/GPT - I failed to mentioned that on the existential question, BARD replied with 3 drafts whereas ChatGPT gave one. In a way BARD gives neutral, positional, and contrary responses and you get to read all of them. I like that.

BARD/GPT - comparison 3/3

What FDA Clinical Trials included NK cells?

BARD - There are many clinical trials that have included NK cells. Here are a few examples:

A phase I trial of COH06, an allogeneic NK cell therapy, in patients with non-small cell lung cancer previously treated with PD-1 and/or PD-L1 immune checkpoint inhibitors that has spread to other places in the body (advanced) and that has not responded to previous treatment (refractory).
A phase I trial of CNK-001, a cryopreserved human placental hematopoietic stem cell-derived natural killer (NK) cell therapy that is not genetically modified, for the treatment of patients with acute myeloid leukemia (AML).
A phase II trial of NK cells plus nivolumab in patients with advanced melanoma.
A phase II trial of NK cells plus atezolizumab in patients with advanced non-small cell lung cancer.
A phase III trial of NK cells plus atezolizumab in patients with advanced non-small cell lung cancer.
These trials are still ongoing, but the results so far have been encouraging. NK cells have been shown to be safe and well-tolerated in these trials, and they have also shown some evidence of clinical benefit.

Sources - Learn more
clinicaltrials.gov/ct2/show/NCT05334329

ChatGPT

The FDA has approved several clinical trials that involve the use of natural killer (NK) cells. These trials are being conducted for various indications, including cancer, infectious diseases, and autoimmune disorders.

Here are a few examples of FDA-approved clinical trials that include NK cells:

A Phase I/II clinical trial (NCT03671295) is currently ongoing to evaluate the safety and efficacy of using NK cells in combination with chemotherapy for the treatment of relapsed or refractory acute myeloid leukemia (AML).

A Phase I/II clinical trial (NCT03081780) is evaluating the safety and efficacy of using NK cells in combination with monoclonal antibodies for the treatment of relapsed or refractory multiple myeloma.

A Phase II clinical trial (NCT03589339) is evaluating the safety and efficacy of using NK cells in combination with stem cell transplantation for the treatment of high-risk acute leukemia.

A Phase I clinical trial (NCT04107142) is evaluating the safety and efficacy of using NK cells as a treatment for solid tumors.

A Phase I/II clinical trial (NCT03193176) is evaluating the safety and efficacy of using NK cells for the treatment of COVID-19.

It's worth noting that this is not an exhaustive list, and there may be additional clinical trials involving NK cells that have been approved by the FDA.

BARD/GPT - comparison 2/3

How many cells will it take to model Earth's climate with accuracy, extrapolate from current models if necessary?

BARD -
The number of cells needed to model Earth's climate with accuracy depends on the resolution of the model. A higher-resolution model will need more cells to accurately represent the details of the climate system. However, higher-resolution models also require more computing power to run.

Current climate models use grid cells that are typically 100-200 kilometers in size. This resolution is sufficient to capture the large-scale features of the climate system, such as the jet streams and the El Niño Southern Oscillation. However, it is not sufficient to capture the smaller-scale features, such as clouds and precipitation.

To improve the accuracy of climate models, scientists are developing models with higher resolution. The National Center for Atmospheric Research (NCAR) is developing a model with grid cells that are 10 kilometers in size. This model will be able to capture the smaller-scale features of the climate system, which will improve its ability to predict future climate change.

However, even with higher-resolution models, it will not be possible to perfectly model Earth's climate. The climate system is too complex and there are too many factors that affect it. Climate models are simply tools that help us to understand the climate system and to make better decisions about how to respond to climate change.

ChatGPT -
It is not possible to determine the exact number of cells required to model Earth's climate with accuracy as it depends on the complexity of the model and the level of detail required. However, current state-of-the-art global climate models use a grid resolution of approximately 100 kilometers, which translates to millions of cells.

In recent years, there has been a push towards increasing the resolution of climate models, which can require trillions of cells. For example, a study published in the journal Nature in 2019 used a climate model with a resolution of 25 kilometers to simulate extreme weather events in North America. This model required over 3.5 trillion grid cells.

It is worth noting that the accuracy of a climate model does not depend solely on the number of cells used, but also on the quality and quantity of input data, as well as the assumptions and approximations made in the model. Therefore, increasing the number of cells alone may not necessarily lead to a more accurate model.

BARD/GPT - comparison 1/3

What is the difference between Newtonian and Non-Newtonian calculi?

BARD -
Newtonian calculus is the calculus that is based on the work of Isaac Newton in the 17th century. It is based on the idea of derivatives and integrals, which are used to calculate the rates of change and areas under curves.

Non-Newtonian calculus is a family of calculi that are not based on the work of Newton. They are based on other ideas, such as the idea of differential forms or the idea of geometric calculus.

The main difference between Newtonian and Non-Newtonian calculus is that Newtonian calculus is linear, while Non-Newtonian calculus is nonlinear. This means that in Newtonian calculus, the derivative of a function is a linear operator, while in Non-Newtonian calculus, the derivative of a function may be a nonlinear operator.

Non-Newtonian calculus is often used in areas where Newtonian calculus is not sufficient, such as in the study of turbulence or in the study of materials that are not solid or fluid.

ChatGPT -
Newtonian calculus refers to the standard calculus developed by Sir Isaac Newton and Gottfried Wilhelm Leibniz in the 17th century. It deals with the study of rates of change and accumulation of quantities. Newtonian calculus is based on the assumption that functions are continuous, differentiable, and have a constant rate of change.

Non-Newtonian calculi, on the other hand, are a family of alternative calculi that relax one or more of the assumptions of Newtonian calculus. Non-Newtonian calculi have been developed to deal with situations where the assumptions of Newtonian calculus are not satisfied, such as in the case of discontinuous or non-differentiable functions.

There are several types of non-Newtonian calculi, including:

Fractional calculus: This deals with derivatives and integrals of non-integer order, which can be used to describe systems with long-range interactions or memory effects.

Infinitesimal calculus: This deals with infinitesimal quantities that are smaller than any real number, which can be used to describe phenomena that occur on a very small scale.

Ultradifferentiable calculus: This deals with functions that are infinitely differentiable but not necessarily analytic, which can be used to describe functions with singularities or rapid growth.

Quantum calculus: This deals with operators and functions that behave like quantum mechanical operators, which can be used to describe systems in quantum mechanics.

In summary, while Newtonian calculus assumes continuity, differentiability, and a constant rate of change, non-Newtonian calculi relax one or more of these assumptions to deal with more complex systems.

Received the waitlist invitation email at 8:40am. Offer to try it out at 2:50pm. I was not the only one in a rush I guess.

I ran 4 comparisons between Bard and ChatGPT. I will post 3 of them. The 4th one dealt with religions and cults and prob too controversial here. I noticed that it took ChatGPT longer (>10X) on science/tech questions.

Comparisons to follow in next 3 posts.

GOOG Bard - tried it yesterday but was in a rush. First impression is that it is much better than the first release of ChatGPT I will do a comparison later in the day with latest ChatGPT.

GOOG Bard - Sounds like the legal department told them to assume some users are dumb. LOL.

I am on the waitlist BTW and that means just about everybody is as well.

The implant process had to be re-designed for whatever reason. This was also cited in some lawsuits by clinics.

Haven't followed for a while and saw the PR today. The first version didn't work as advertised and that's a big problem for any surgical robot. Maybe this one actually works but I will just watch with curiosity for now. I don't understand the "bail in" of Restoration Robotics given the circumstances.

AAPL EV? Not so fast. From Bloomberg

"Foxconn is finding it’s much harder to build electric cars than iPhones and tablets, Sean O’Kane writes for Bloomberg Businessweek. .............. So far, it’s only made a handful of prototypes, a few dozen electric buses and about 40 pickups for Lordstown Motors, which in January asked Foxconn to suspend production because the cost of making its trucks exceeded the targeted sale price of $65,000."

I couldn't get past the top of the list and only glanced at the rest. I recall some prominent ones were missing. FATE, TAK, NKTX so prob no.

Thx. Noticeable improvement in accuracy. Re last item on cancer research - One question I posed asked "what NK cell therapies for cancer are currently in clinical trials?" in the last ChatGPT version, it came back with a list topped by ADAP's MAGE-A4 TCR. At least it knows it has something to do with cell Tx.

GPT 1 July 2018
GPT 2 Feb 2019
GPT 3 June 2020
ChatGPT Nov 2022
ChatGPT.2 March 2023

Software doesn't scale like Moore's law so expect cycles to get shorter with each release but I have no idea how usefulness and accuracy will be measured. Some leading AI researchers already find it useful for debugging and writing code.

GPT4 - Up and running in ChatGPT. Haven't tried it yet. Maybe they should come up with an application specific version to replace humans at banks that don't know how to manage deposits.