AI just changed the world for real. We all need to understand what just happened.

Jumping right in - I'll address several things I've been asked a lot about lately just to clear the air, and then we need to jump right in to the recent AI news updates such as MultiOn's Agent Q, AI Scientist, Cosine Genie, SearchGPT and other new and rumored OpenAI updates, Google Voice, "Prompt Caching" by Anthropic and more, and especially how it pertains to your life (how you can leverage it, what to be cognizant of, et al).

First off - I am not @iruletheworldmo ?????? on X nor do I know much about him/her/them despite the account following me and engaging with me there. There has been speculation (even sparked by the account) that this is Sam Altman's alt-account, the team at MultiOn, and even an AI agent, but I can neither confirm nor deny any of that, other than congratulating the account for growing to 35K followers there in weeks, for the viral tweets and engagement from top AI accounts (even and including SamA), and also must acknowledge that not everything this person ('Mo'?) has forecasted has come to fruition (but some has). A second account, Lily Ashwood (yes, it says "Ilya" in the middle) @lilyofashwood (X) has been alledged to be an AI agent who's posting, hosting live Twitter Spaces and more. Regardless of whether "she" is a real person or an AI, we definitely do need to start preparing for a future soon where a solid portion of engagement and accounts on social media are not "real people" (humans), but rather AI bots.

Secondly, I will NOT be giving much visibility to or discussing xAI and 'Grok' much at all, so will simply say that in my vicarious experience that the only thing Grok is really "great" (or even very good) at is its image generation - BUT there are several things about that that need to be known at scale, so I will simply touch on that, namely that that image generator is unhinged, but also for free outside of Grok, so you don't need a subscription to Twitter to access it. I know many of you want to hear more about my feelings on the company and model, but I have many reasons to respectfully decline to include x/elon in this newsletter to the degree I will highlight OpenAI, Anthropic, Google, Mistral, Meta and other market leaders.

Thirdly, and finally (I'll provide commentary as I discuss the new AI capabilities and news below), I am hosting a FREE masterclass with AI ethics expert Ravit Dotan, PhD on "How To Build Using AI, Ethically) on 9/9 at 12 noon EST - info and registration (all free) here: https://www.dhirubhai.net/posts/corywarfield_activity-7230328586103119873-TGar and feel free to let friends know as well; we will go over how to build apps/websites/business plans using AI, and what to look for to ensure alignment, ethics, and ease.

So now, without further ado, is a wholly mediocre "Blockbuster Movie Poster for Tech For Good Linkedin Newsletter by Cory Lopes-Warfield" made by Flux1, the free, open-source and although it's embarrassing, I wanted to share the image it created of me, which looks like I'm a tardigrade auditioning for a reboot of "Goonies".

So now, AI NEWS that really matters! Welcome to Tech For Good Issue #13!

So, am I being dramatic? Did AI really just change the world? Well, I don't believe that I'm being overtly dramatic or inflating what's really begun being released to the wild, and the implications, at least in my assessment, are huge.

So to get right into it, we all need to know about Sakana AI out of Japan, and thie new "AI Scientist" that can conduct research autonomously, but also run experiments, come up with novel research concepts, visualize and present, write full academic research papers, and perform peer review. This can and likely will lead to accelerated research, increased productivity, new innovations, treatments, actionable insights, democratized research and much more, but we must be mindful of biases, quality control, and the potential job displacement that systems like this will almost certainly result in. We continue to see more, more capable agentic and autonomous agent systems emerge, even as Claude, ChatGPT, Microsoft and others build and launch their own, but this AI Scientist seems to be in a class of its own, at least thusly, and the implications are profound (to touch on the job displacement, this is why Universal Basic Compute, or Universal Basic Income is a conversation that needs to be perpetuated, at scale, and continue to be a central societal conversation). This innovation is a huge step closer to what is called the Intelligence Explosion, which may be a step toward "singularity".

In other AI Agents news, MultiOn , who I've been covering since my "AI Logs" newsletter days, just released "Agent Q", which is also a highly capable technology deploying autonomous AI agents that have been proven out in real-world use-cases. In one such recent example, Agent Q was able to make real-world bookings via OpenTable 64% more efficient and reliable (from 18.6% to 81.7%) using LLaMA-3. OpenTable is also partnered with OpenAI , and are emerging as a truly forward-thinking tech company.

I've showcased many AI's capable in coding, some more competent than others, over the past year of this newsletter and the previous "AI News", but until now, the forerunner had been Cognition Labs' Devin (recently received a major investment from 微软 ). As of right now, a new AI programming platform has emerged as the new industry leading benchmark (performing twice as good as Devin, and many times better than ChatGPT, Gemini, even Claude) called Genie (by Cosine ) - I've featured them before too, but their newest release is an order of magnitude better, and truly newsworthy. The fact that AI is getting this good at writing, testing, documenting, deploying, and updating code and software is truly mindblowing.

Now, image a "mixture of experts" using the best in class AIs interchangeably as autonomous agents (Agent Q) with the sophistication and capability of a 'AI Scientist" that has access to programming such as Genie (or even Devin). NOW imagine this powered by an even smarter and more capable LLM (possibly GPT5), and much more notably now imagine all this embodied in a robot that can see, act, move, assess, build, assist, and much more. We will not end this year without seeing a mass deployment of humanoid robots (androids) powered by this powerful combination of technology.

Is that all?

No, not at all. OpenAI released a new model last week inside ChatGPT, new potential OpenAI models are surfacing on LMSYS (large model systems arena to blind test new models), as well as a new paper and disrupted a covert Iranian influence operation to influence the US elections (a subject I'm particularly sensitive to), but the big update was SearchGPT, which I have been provided early access to. My initial assessment is that I love it as it's significantly better than Googe search's 'summaries', cites its sources clearly, presents a unified answer to my queries with no ads, and is an overall better search experience than Google, Bing, and even Perplexity yet (in my own opinion). If you haven't applied for early access, I recommend doing so, otherwise I believe it will be publicly available very soon. If you've also tried it and have a reaction to it, please share it in the comments of this newsletter post - I'd love to hear your experiences with it.

The AI race is clearly* between OpenAI and Google (although that could be read Microsoft AI and Google, or at least could be soon), and OpenAI has successfully stolen Google's thunder several times in the past year when it comes to announcements and releases, but this time Google was able to emerge victorious in the race to launch "Voice Tech" (which they both licensed from Microsoft AI Co-Innovation Lab when they received exclusive licensing rights to all Inflection AI tech in their "non-hostile" "non-takeover" of Inflection several months back, days before OpenAI and Google both demoed "their new" voice AI technologies. Google Voice had some hiccups in their live demo, but to their credit its begun rolling out and the responses are relatively positive. It really is like having a conversation with a colleague, mentor, friend, and helps many people seemingly get better results out of using AI, which feels net positive.

I put an asterisk next to 'clearly' above because many will argue that elon's xAI is now a top competitor as well, after having raised a ton of money, launching "Grok2" to fanfare, and having it's reception prior to launch on LMSYS as "suscolumn-r" be said to be comparable to ChatGPT, Gemini, and Claude, but my assessment is that it's really not as good in my opinion and experience as the others - all of whom are about to release their next-gen models - and the selling point is it's photorealistic image generator with no filters except nudity, but again, this is not Grok, it's simply them embedding Flux1 AI image generator, and removing all restrictions so that it can be used on X to create deepfakes, disinformation, violate trademarks blatently, and otherwise cause rift and division. Therefore, I personally do not consider them a top competitor, and also I do not support elon musk, neuralink, tesla/bots, X or starlink. Obviously all reader are very welcome to hold and honor their own opinions, but I reserve the right to share mine freely as well.

Finally, I've mentioned that a startup I've cofounded is helping teach computers and AI "what love is" and why it matters, how to leverage it for better outcomes as we continue to rush head-on into this new "intelligent world" and keep it "on the rails" here previously, and the amount of DMs, comments and interest I've received around it already has been overwhelming. We are building somewhat in shadow mode with an ambitious roadmap, but if you're curious how to get involved, collaborate or learn more, feel free to comment or shoot me a DM as well. I believe this could be a huge component of the winning tech-stack for humanity.

That's all folks! I hope to see you at the free masterclass on how to build using AI with Ravit on 9/9 (link above). I am also teaching a linkedin viral growth course soon with Sofia Wakabayashi or WakaGrowth , a Linkedin Live with Guidde on AI Customer Onboarding soon, am "Building In Public" every week with WeJ - Where Everyone's Jamming where we invite YOU and everyone to be flies on the walls on our weekly executive standup meetings where we discuss our tech, how we're using other tech to build with, etc.

Oh and I can't omit that on the 27th, I will be interviewing Presidential Candidate Robert F. Kennedy Jr.(RFK Jr.) here LIVE on LinkedIn on the 27th of this month (August) where we will discuss AI, the future of work, blockchain and many matters #techforgood related - I welcome all to join us for that as well, and I'm in talks with the campaign to do a live with his running mate Nicole Shanahan here very soon as well.

Thank you for taking the time to read this 13th issue of Tech For Good newsletter. If you enjoyed it and aren't yet subscribed, I welcome you to do so (it's free and there's no spam); if you already are - why not share it with a friend? The more we all know, the more we all know! TOGETHER WE RISE! I'm grateful for the 170,000 subscribers this publication has amassed already inside it's first 4 months - I couldn't (and wouldn't) do it without all of you!! So a massive "thank you" from my heart, and wishing all a great weekend and week ahead. Stay tuned - I have it on authority that this week will be another big one for the world and the world of Artificial Intelligence. May we all remain well and good, even amidst a intelligence explosion.

Signing off,

Cory Warfield , Coryconnects .