Thinking Machines Lab is ex=
-OpenAI CTO Mira Murati's new startup (3 minute read)
OpenAI's former CTO Mira Murati's new s=
tartup, Thinking Machines Lab, has come out of stealth. The startup intends=
to build tooling to make AI work for peoples' unique needs and goals and t=
o create AI systems that are more widely understood, customizable, and gene=
rality capable than those currently available. It will focus on building mu=
ltimodal systems that can adapt to the full spectrum of human expertise and=
enable broader applications. Murati joins a growing list of former OpenAI =
execs who have launched startups.
|
OpenAI board considers spec=
ial voting powers to prevent Elon Musk takeover (2 minute read)
OpenAI is considering granting new voti=
ng rights to its nonprofit board to help it fight the unsolicited takeover =
bid from Elon Musk. Giving the nonprofit's board outsized voting power woul=
d ensure it retains control of the restructured company and is able to over=
rule other investors, ensuring OpenAI can fight off hostile bids from outsi=
ders. OpenAI could also consider a plan that would let shareholders buy up =
additional shares at a discount to fend off hostile takeovers.
|
|
Science & Futuristic Technology
|
Scientists Created the Ligh=
test and Strongest Nanomaterial Ever (3 minute read)
Researchers at the University of Toront=
o have created a new material with the toughness of steel that weighs about=
as much as foam. They used AI to recognize the best configurations of nano=
structures to create the tough and impossibly light material. The new nanom=
aterial can be mass-produced, unlike its predecessors. The lightweight natu=
re of the material could result in more comfortable prosthetics and implant=
s and more efficient vehicles.
|
AI used to design a multi-s=
tep enzyme that can digest some plastics (4 minute read)
AI-driven protein design has enabled th=
e possibility of creating things unlike anything found in nature. Scientist=
s have successfully created an enzyme with the potential to digest plastics=
.. Breaking down ester bonds in plastics requires four steps - getting AI to=
design a protein with the right configuration to do one of these steps is =
easy, but having it cycle through all four is much harder. The scientists o=
vercame challenges by adding more AI models and eventually designed an este=
rase capable of digesting the bonds in PET, a common plastic.
|
|
Programming, Design & Data Science
|
My LLM codegen workflow atm=
(15 minute read)
This article discusses a workflow that =
uses large language models for code generation (LLM codegen) that has been =
effective for building many small products. The workflow involves brainstor=
ming a spec, making a plan, and then executing that plan using LLM codegen.=
The process is repeated until magic happens.
|
Kafka at the low end: how b=
ad can it get? (3 minute read)
Kafka does poorly as a job queue. It un=
fairly assigns jobs to workers, even if other workers have nothing to do. K=
afka is not a bad tool, it is just not designed for low volume - it was des=
igned for scenarios where a conventional single-node message broker can't k=
eep up. If that speed isn't required, then it's better to use a conventiona=
l broker with more suitable features (at least until Queues for Kafka comes=
along).
|
|
The One AI to Rule Them All=
(8 minute read)
The next AI model war may be upon us, b=
ut this time it will be a smaller one as the gains, while real, have been i=
ncreasingly expensive and granular. We are seemingly reaching parity among =
the players. While they all have more specific areas of focus and subfeatur=
es, they are all trying to consolidate their products to simplify them for =
everyday use. This article ranks current AI offerings from a pure consumer =
perspective.
|
The marketing genius of Bry=
an Johnson (4 minute read)
Bryan Johnson's Don't Die lifestyle has=
grabbed the world's attention. Johnson stands out - no one else is making =
content about not dying. His 'Don't Die' slogan is hard to scroll past, sho=
rt, and snappy. Johnson doesn't try to appeal to everyone - his extreme lif=
estyle polarizes people, but he only needs a small number of fanatical fans=
.. This article looks at Johnson's marketing tactics, which have resulted in=
1,440,000 YouTube subscribers, 503,000 followers on X, and 1,500,000 Insta=
gram followers.
|
TLDR is hiring curators for=
our new TLDR Data newsletter (Fully Remote, $100/hr)
TLDR is hiring part time curators to la=
unch our TLDR Data newsletter.The ideal candidate would have deep=
experience working directly with data warehouses, data pipelines, data lak=
es, and other modern cloud infrastructure. Time commitment is ~2-3 hours/week paid at a rate of $100/hr. To apply please send y=
our LinkedIn or resume to jobs@tldr.tech alo=
ng with a couple sentences on why you'd be a good fit!
|
|
|
Love TLDR? Tell your friends and get rewards!
|
|
Share your referral link below with friends to get free TLDR swag!
|
|
|
|
|
Track your referrals here.
|
|
Want to advertise in TLDR? =F0=9F=93=B0
If your company is interested in reaching an audience of tech executives, d=
ecision-makers and engineers, you may want to advertise with us.
Want to work at TLDR? =F0=9F=92=BC
Apply here or send a friend's resume to jobs@tldr.tech and get $1k if we hire them!
If you have any comments or feedback, just respond to this email!
Thanks for reading,
Dan Ni &a=
mp; Stephen =
Flanders
|
|
|
|