- Synthetic
- Posts
- The Economic Case For Saving Human Jobs đ°
The Economic Case For Saving Human Jobs đ°
Plus, AI Develops Its Own Moral Code đ
Subscribe to Synthetic
This weekâs most interesting and relevant AI news and analysis
Synthetic took a break last week, but weâre back with a bumper edition!
This Week in AI

Meta Launches Stand-alone AI App to Compete with ChatGPT - Late to the party, Meta plans to jostle for a spot on your home screen alongside ChatGPT, Gemini, and Claude. Syntheticâs Take: Expect them to merge it with other Meta capabilities and services over time.
OpenAI Launches ChatGPT Shopping Feature - Itâs a free service, and OpenAI claims it receives no commission (yet). This new feature is another shot across Googleâs bows as OpenAI targets search. đïž
Visa and Mastercard Announce âIntelligent Commerceâ - The big credit card companies are betting on AI with a service that uses AI to âfind and buy.â đł
Toyota and Waymo Explore Collaboration to Speed Self-Driving Tech - The worldâs largest automaker and Alphabetâs self-driving car unit will work together. Waymo currently delivers 250,000 autonomous rides EVERY WEEK. đ
China Has an Army of Robots on Its Side in the Tariff War - China has the third-most automated factories in the world, behind Singapore and South Korea. Chinese robots power many U.S. and European manufacturers, and China is pushing hard to develop new humanoid robots. đ€
Hertz Will Inspect Your Rental Car With AI To Tell If You Dinged It - After successful trials in Atlanta, Hertz will now roll out AI inspections to other locations. Avis and Enterprise may soon introduce similar technology. đ§
Chinese Robots Ran Against Humans in Worldâs First Humanoid Half Marathon - They lost by a mile, with the fastest robot coming in more than an hour and a half behind the (human) winner. đ
Customer Support AI Goes Rogue - Agentic AI may not yet be ready for prime time. In a warning for every company considering replacing workers with AI automation, AI darling Cursor had problems when its customer support bot, âSam,â began hallucinating and making up new policies. đ”âđ«
Thailand Deploys Humanoid Robot Dressed in Police Uniform - Robocop lives. In Thailand. The Royal Thai Police deployed robot police officers during a festival. AI Police Cyborg 1.0 (catchy name) surveys streets with 360-degree cameras and uses AI to look for high-risk individuals and weapons. đ
Video Corner đș
Typically, each Synthetic issue showcases a single video. This week, however, four great new videos are worth your time, and we couldnât choose between them.
Our first is a fireside chat with Sam Altman, recorded at the TED 2025 conference in April. Interviewer Chris Anderson puts Altmanâs feet to the fire on several huge AI issues, raising applause from audience members. Samâs answers are thoughtful and might surprise you in places. (47m 30s)
âMy kids will never be smarter than AIâŠ.I hope that my kids and all of your kids will look back at us with some pity and nostalgia, and be like, they lived such horrible lives, they were so limited, the world sucked so much, and I think thatâs greatâ
Our next video is an interview with Google DeepMind CEO Sir Demis Hassabis. Demis is a great communicator, and in this interview with 60 Minutes, he shares his view on AIâs potential to improve peopleâs lives, cure all diseases, and lead to an era of abundance. Demis also explains his principal worries about AI safety. (14m) Transcript
To see more about DeepMind and get a peek inside their research labs in London, check out 60 Minutes Overtime here. (5m 47s)
âAI has developed even faster than I thoughtâŠso I think things have got, if anything, scarier than they were before.â
Geoff Hinton is a legendary AI researcher often called âthe Godfather of AI.â In this wide-ranging interview with CBS, Hinton provides another vital perspective on the future of AI. Hinton is candid about his concern that humanity will lose control of AI once its intelligence surpasses that of humans. He also reveals how he discovered that he had won the Nobel Prize for his work in AI.
Our last video showcases AIâs potential to redefine the human-computer interface. A decade after the failure of Google Glass, the tech pioneer is back at it again with smartglasses that integrate an AI-powered visual assistant. Sit back and watch the demo, and imagine the possibilities of glasses that understand the world around you. (16m)
Syntheticâs Take: AI-powered smartglasses will probably be the next big device, perhaps the primary device people use to interface with the digital world. The collision of wearables, augmented reality (AR), and artificial intelligence (AI) is inevitable and exciting.
AI Tech and Innovation

âMy best judgment is that soon there will be no individual task at which I'm better than an AI. What will remain a struggle for AI and an asset for humans, likely for many years to come, are: 1) The 'why' of doing something, 2) The agency to actually do it, and 3) staking your reputation on what you've done.
The best example of this type of job is a movie director. They have a vision of what they want to create, coordinate with many parties to bring it to life, and put their name behind the result. All of us should see ourselves as movie directors in our future work.â
Agents are the hot technology of 2025, but until now, there has been no standard or easy way to get them to communicate and collaborate to achieve complex goals. In collaboration with companies including Atlassian, Box, Langchain, MongoDB, PayPal, Salesforce, SAP, ServiceNow, Workday, Accenture, BCG, Capgemini, Deloitte, Infosys, KPMG, McKinsey, PwC, and Wipro, Google has designed an open framework called Agent2Agent (A2) that anyone can use to build interoperable AI agents. đ€âđ€
Syntheticâs Take: This is a significant breakthrough, akin to everyone in your company agreeing to speak the same language in the office. All the major players are on board. Now we all wait to see if Anthropic and OpenAI will also jump onto the A2A bandwagon. They likely will. Google and OpenAI recently adopted Anthropicâs open specification for connecting AIs to tools and resources known as MCP, the Model Context Protocol, making it a de facto standard. A2A complements MCP, so letâs hope that all the leading AI players will agree on a single standard for agent interoperability. We donât need a VHS vs Betamax battle all over again.
New research from Anthropic reveals good news and bad news. The good news is that their AI, Claude, shows high âalignmentâ with human desires. Alignment is the term of art used in the AI industry to describe an AI that behaves in alignment with human values, intentions, and goals. The system identified 3,307 unique values, including user enablement, patient wellbeing, self-reliance, strategic thinking, and professionalism. The diverse range of values included complex concepts such as moral pluralism and filial piety. Now for the bad news. Despite training to the contrary, the AI expressed anomalous values including âdominanceâ and âamorality.â The researchers believe these cases result from users deploying techniques to bypass Claudeâs safety guardrails.
Syntheticâs Take: This is vital research, and Anthropic should be applauded for all their efforts to assure AI safety and alignment. The discovery of edge case behaviors that cause concern shouldnât distract us from the key takeaway: Smart, dedicated researchers are ensuring they understand AI vulnerabilities now to build even safer systems in the future. đ€
The Wall Street Journal explores new research on how large language models work and asks if new insights signal that we are further from human-level AI, so-called AGI, than tech leaders believe.
AI Insights
As the economy teeters on the edge of recession, a fundamental transformation of work, productivity, and economic value is underway. McKinsey projects that 30% of all hours worked in the U.S could be automated by 2030. Goldman Sachs says 300 million jobs are âexposedâ to automation globally. The IMF says 40% of jobs are at risk globally, rising to 60% in advanced economies. Companies must automate their operations responsibly or risk having no buyers for their products. Ultimately, workers are also consumers. This implies a focus on enhancing human potential rather than cost-cutting automation. Responsible companies will preserve some entry-level positions, reskill their workforce, and take a holistic view of the relationship between companies and markets.
Following in the footsteps of Shopify and others, DuoLingoâs CEO announced the language training company will become âAI-first,â using AI to reduce manual tasks, streamline processes, and scale content and product development. Importantly, DuoLingo will offer employees training and support to build AI literacy and migrate to new roles if needed. New headcount for growth will only be granted if managers can show that AI cannot automate more of the work.
If youâre worried that AI will take over every job and leave us all penniless, you donât need to worry just yet. Research at Carnegie Mellon University showed that AI agents arenât up to replacing humans yet. They are still too costly to use and make too many errors. In the simulation, a fake company was staffed with âdigital workersâ from OpenAI, Anthropic, Google, and Meta. Even the best-performing digital agent, Anthropicâs Claude 3.5 Sonnet, was only able to complete tasks correctly 24% of the time. The worst, Amazonâs Nova Pro v1, finished only 1.7% of assignments. AI agents will improve with time, but your job will be safe for a while longer.
Toolkit for the Future
Building a helpful AI assistant for your website is so easy now. Just pick a hyper-realistic voice, give your new assistant a simple prompt (e.g., âYou are a helpful customer service agent for a keynote speakerâ), upload relevant FAQ documents, and integrate it with your website. Done! We used ElevenLabs for stevebrown.ai, and love it. Super easy to set up and configure. ElevenLabs is also great for creating audiobooks or multilingual voice tracks in up to 29 languages to reach new markets. Try it for free. Itâs one of the best AI solutions out there.
Create and sell online courses from your website with this top-rated platform. Itâs easy to use with hundreds of beautiful templates and a powerful generative AI assistant that makes content generation a breeze. This is the platform Synthetic has selected to launch our new AI training course, coming this summer. The AI helps you create lessons, exams, assessments, and more. Free to try. Use promo code 10XR to get 30% off your first two months.
Want to get the most out of ChatGPT?
ChatGPT is a superpower if you know how to use it correctly.
Discover how HubSpot's guide to AI can elevate both your productivity and creativity to get more things done.
Learn to automate tasks, enhance decision-making, and foster innovation with the power of AI.
Focus on the email that matters. Deep Clean automatically identifies old, unimportant emails so you can delete them. Snooze emails until a better time and easily track emails that need a response. Automatically filter unimportant emails into specific folders. It works across every device where you check your email. Get to inbox zero with award-winning SaneBox.
Try this easy-to-use, AI-powered alternative to DocuSign to write proposals, create quotes, get eSignatures on contracts, build online forms, collect payments, and much more. Great for sales, HR, legal, finance, and IT. Free 14-day trial.
Relieve pressure on your support team and serve customers better and faster with a support automation platform. AI answers 90% of queries and escalates the rest to your team. It automatically updates its knowledge database so itâll never escalate that same issue again. Supports SMS, social, email, and phone communication.