An excellent overview of Large Language Models, LLMs, & what are they used for from @nvidia@NVIDIAAI :
blogs.nvidia.com/blog/2023/0…
LLMs are making Artificial Intelligence more accessible to the end user & have placed AI on the radar of many CEOs. They have even been embedded into robotics enabling humans & machines to converse with each other.
LLMs are also known as foundational models based on Transformers with Self-Attention Mechanism introduced by Google researchers in 2017 and the next generation is set to be increasingly multimodal working with text, speech, images and video. They are also advancing on reasoning skills.
The State-of-the-art models have Reinforcement Learning as a reward based mechanism to help deliver the correct result and increasingly use human feedback.
There is also a growing base of open source models available alongside the proprietary models from Open AI & Google.
@pierrepinna| @sallyeaves | @SpirosMargaris | @TheRudinGroup | @HaroldSinnott | @Nicochan33 | @Ym78200 | @lyakovet | @gvalan | @Shi4Tech | @BetaMoroney |@ahier | @mikeflache| @ipfconline1#AI#MachineLearning#LLMs#Fintech#DeepLearning#legaltech
An excellent summary on the era of Artificial Intelligence - Why AI will save the world:
a16z.com/2023/06/06/ai-will-…
by @pmarca |@a16z
An important point that is raised big AI companies should not be "allowed to achieve regulatory capture, not allowed to establish a government-protect cartel that is insulated from market competition due to incorrect claims of AI risk."
IMHO we will still end up with some form of AI regulation but it should not be designed to protect big players at the expense of startups & other new entrants using Open Source.
@vonderleyen@POTUS@RishiSunak@Keir_Starmer@vestager@VanberghenEU
AI can be used to great effect to help advance healthcare and the fight against climate change.
Regulation should be aimed at making things better for the user (our citizens) and society and economy (GDP, living standards) and improving competition - not worsening it as that results in worse outcomes for both the individual and society overall.
Any regulation needs to take account of the access to data and probably revamp privacy rules such as GDPR & align competition law accordingly.
Policy makers should mandate data trusts that would protect the end user & also enable new entrant startups to access data.
They should also create Generally Acceptable Privacy Policy (GAPP) and Generally Acceptable AI Development Policy (GAAIDP) as principle based standards that may be auditable for production models.
Policy makers should also encourage collaboration on synthetic data and federated learning with differential privacy so that we can find ways to effectively scale AI across all sectors of the economy whilst protecting user privacy and also encouraging competition.
@pierrepinna| @sallyeaves | @SpirosMargaris | @HaroldSinnott |@TheRudinGroup | @Nicochan33 | @gvalan | @mikeflache | @ahier| @lyakovet | @BetaMoroney | @Shi4Tech |@Ym78200| @ipfconline1 | @pierrecappelli | @pawlowskimario#AI#MachineLearning#DeepLearning#LLM#startups#ESG
Turkish bank shares were also hit hard, with the country’s main banking index falling by more than 8%.
The lira was at 19.65 to the dollar on Monday morning, its weakest since a record low of 19.80 in March.
guardian, TR debt already solved though rest in the way
#Fintechs and incumbent #banks are turning to generative AI and LLMs—powered by AWS and NVIDIA platforms—to enhance customer experience including chatbots, digital avatars, hyper-personalization, and more. Read the TechCrunch article. bit.ly/3LTnDHv