Meta GPT3 Rival + ChatGPT Recovers $100K

Facebooks GPT3 Rival

Mark Zuckerberg is at it again, and this time, he's not “metaversing” around.

On Friday, Meta dropped a bombshell and announced the release of their new language model called LLaMA, and let me tell you, it's not your average run-of-the-mill language model.

LLaMA comes in different sizes, just like a pair of shoes. You can choose from four sizes: 7B, 13B, 33B, and 65B parameters. Sure, it may be smaller than GPT-3 (175B), but LLaMA is specifically designed to complete more specific tasks like a boss.

What's more, this language model packs a punch. Even with over 10 times fewer parameters, LLaMA's 13B version still managed to outperform GPT-3's 175B version in some tasks. And that's not all; it also beat out Meta's previous language model OPT and gave Google's PaLM a run for its money. LLaMA is like that small but mighty friend you always need on your side.

But wait, there's more! Meta has pledged to make LLaMA open to the public. Unlike OpenAI, who's faced criticism for not being truly "open," Meta is sharing the love and plans to release their model and share the code with other researchers. Researchers will get first dibs on LLaMA, and it's going to be a bit exclusive. Meta plans to limit access to academic researchers and government organizations initially. They're doing this because researchers often face difficulties accessing LLMs due to the high costs involved in running them.

Read More

CEO uses ChatGPT to Recover $109,500

A CEO used ChatGPT to recover $109,500 from a client, here’s how he did it

Here’s the full thread: https://twitter.com/gregisenberg/status/1629158318915653633

Here’s the thread breakdown:

A US-based CEO recently shared how he skipped the expensive legal route and used ChatGPT to warn a client who went radio silent on payments.

Isenberg said he instructed ChatGPT to imagine itself as a finance department employee and draft a "scary" mail asking the client to pay up immediately.

It went on to remind the client how their behavior would damage their standing and invite legal action, adding that they must clear their dues within three business days.

While Isenberg was glad he didn't have to hire a lawyer, he said the episode got him thinking about how much work could be automated with AI.

Scammers are spoofing ChatGPT to spread malware

ChatGPT is the new kid on the block and already gaining major popularity as the fastest growing app of all time. But where there are users, there are hackers. Yup, these sneaky cyber criminals are trying to scam innocent internet surfers by setting up fake ChatGPT websites that spread malware and steal private information. One website, "chat-gpt-pc.online," tried to convince visitors that ChatGPT could be downloaded as a Windows application. However, this download injects RedLine malware into your computer that can steal your sensitive information. And if that wasn't bad enough, there are over 50 fake ChatGPT apps in the Google Play Store that deploy phishing campaigns to steal users' information. One such app, "chatGPT1," is essentially a fraud that secretly subscribes its target to numerous paid services, also known as SMS billing fraud. Yikes! To avoid getting scammed, go directly to the OpenAI website or AI.com to use ChatGPT safely. So, keep your eyes peeled, folks!

Read More

Monday’s Question of The Day

How do you use AI to boost your daily efficiency?

AI Could be ChatGPT, Dall E, MJ etc.

Reply this Email

Copy/paste the prompt here: https://pastebin.com/gjk9t0jL

Copy/paste the prompt here: https://pastebin.com/BGun7exw

Coatl Spirit of the pink moon scifi Barbie, Dahlia bloomcore spiral, Hans bellmer JeeYoung Lee, Peter mohrbacher --v 4 --s 750